IDASA - Budget Information Service
http://www.idasa.org.za/budgetday
Reviewing Ten
Programmei
Years
of
the
School
Nutrition
Russell Andrew Wildeman and Nobuntu Mbebethoii
December 2005
For more information, contact Russell Wildeman at [email protected] (021-467
5653) or Nobuntu Mbebetho at [email protected] (021-467 5664)
EXECUTIVE SUMMARY
The provision of basic education in favourable social and economic circumstances is a
complex matter. This degree of complexity is increased when basic education must be
delivered to individuals who struggle with the additional burden of poverty. Poverty means
that many young children are food-deprived and would therefore not be able to participate
fully in their own educational development. This is the context of the school nutrition
programme, which spans the entire duration of the post-1994 political and educational
landscape. Since its inception, the school nutrition programme operated as an in-kind transfer
or benefit that was available to the poorest primary school learners. Earlier studies of this
programme argued strongly for the involvement of local communities in the delivery of
school feeding to children. It was argued that an exclusive focus on government-sponsored
delivery mechanisms would stifle the development of nutrition-conscious communities and
reduce the effectiveness of spending on this programme. Despite the urgency of such calls,
the school nutrition programme has not yet spawned broad community participation that
would transform it from an exclusively school feeding programme to a more comprehensive
nutrition programme.
South African researchers are divided about the policy value of the school nutrition
programme in its past and present guises. There are roughly speaking two “camps”: those
who would like to limit the scope and size of the school nutrition programme and those who
would like to see an expanded programme. The former regard the school nutrition programme
as an exclusively feeding programme with insufficient fiscal space for other vital aspects of
an integrated nutrition strategy. This lack of balance between school feeding and broader
socio-medical interventions is argued to reduce the overall effectiveness of the school
nutrition programme. The solution would therefore lie in the containment of school feeding
through narrower targeting and a corresponding up-scaling of socio-medical interventions
such as de-worming and the delivery of micronutrients. There are also positions for the
containment of the school nutrition programme because it is argued that this programme
represents an inadequate embodiment of children’s rights to basic food. This view argues that
the right to basic food must be located elsewhere and supports the restriction of the
programme to smaller numbers of beneficiaries. Those who argue for the expansion of the
school nutrition programme indicate that as long as the present poverty and unemployment
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
conditions prevail, school feeding would remain necessary. Although there has been no actual
demonstration of the cost-effectiveness and cost-efficiency of school feeding, researchers
believe that programmes that have preventative value should be supported, irrespective of the
costs. Cabinet’s support for the expansion of the school nutrition programme appears to have
settled this debate in favour of those who desire further growth and expansion. We argue that
the continued absence of an appropriate balance between school feeding and other nutritional
considerations places the school nutrition programme in the same domain as other policies
that aim to improve the conditions of poor learners. This would de-legitimise narrow targeting
and strengthen arguments for the expansion of the school nutrition programme.
Tremendous changes accompanied the evolution of the school nutrition programme from a
Presidential Lead Project in 1994 to a stand-alone conditional grant administered by the
Department of Education in 2004. The school nutrition programme’s first budget and fiscal
context was the period immediately after 1994 when government actively combated run-away
consumption expenditure. The erstwhile nationalist government was unable to stem the tide of
excessive government consumption expenditure in spite of policy attempts through the White
Paper on Privatisation and Deregulation in the Republic of South Africa 1987. Thus, while
the post-1994 government had to confront the task of development and re-distribution, it
defined the stabilisation of broader macro-economic aggregates as a vital pre-condition for
the former. Both the developmental impetus and the macro-economic imperatives were
defined in the White Paper on Reconstruction and Development 1994. So while a school
nutrition programme was established that would target poor primary school learners, the costs
of delivering this programme in provinces had to avoid an escalation of overall debt and
further government consumption expenditure burdens. This was achieved through the nonexpansion of staff levels for designated RDP projects and budget allocations that did not keep
pace with inflation. We find in this period the roots of the present conditional grant
framework, which still does not conceptualise adequate staff as a necessary pre-condition for
the effective and efficient implementation of grant funding.
The RDP funds that financed designated projects consisted of international donor aid as well
as funds that were creamed off departmental budgets. Departments could in theory access
these resources but had to re-prioritise expenditure in line with government's overall goals.
Amidst widespread concern about the lack of re-prioritisation and the realisation that
departments used RDP funding to avoid re-prioritisation, the RDP Office was closed in 1996.
RDP projects were integrated under the relevant departments and this meant that the next
destination for the primary school nutrition programme was the Department of Health. This
arrangement was effective from 1998/99 until the end of 2003/04, after which the Department
of Education took over responsibility for this programme. Two distinct periods in the
evolution of public finances in South Africa are contained during this period, namely the
austere 1996-2000 Gear period and the upswing in government expenditure after 2000. The
evidence that we present in this paper suggests that while consolidated health spending
received a shot in the arm after 2000, the school nutrition grant suffered large real declines in
the pre-2000 period and in the post-2000 period. This meant that since the inception of the
programme in 1994, it did not experience real positive growth over an eight-year period,
raising serious questions about the political and policy prioritisation of this programme.
However, while real budgeted allocations were still in reverse, more consistent actual
spending took place since 2002. Two explanations can be used to account for improved actual
spending ratios, namely greater experience with this grant given its eight-year history, and/or
possibly the first fruits of reforms to the conditional grant framework. By the time the
Department of Education was ready to assume chief responsibility for the school nutrition
programme, the latter appeared to have halted its highly variable spending record and the
conditions for the implementation of grant funding were more favourable. However, real
declines in budgeted allocations had not been reversed.
2
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
What budget and fiscal context dominated provincial education funding when the school
nutrition programme was transferred to the education sector in 2004? Provincial education
funding after 2000 prioritised small expenditure categories. In fact, the entire redress agenda
after 2000 is driven by items and expenditure categories that comprise a relatively small part
of budgets. The insertion of the school nutrition programme into this context conforms to this
trend and the school nutrition programme therefore joins in the post-2000 provincial
education funding growth. This is indeed the case because the school nutrition programme is
projected to grow at a real average annual rate of 6.4 per cent. Apart from real growth in
2003, this represents the only sustained real positive growth rate since the inception of the
school nutrition programme in 1994. This real growth was made possible by Cabinet’s
explicit commitment to expanding this programme and revisions to the beneficiary databases
in 2004 and 2005. Another important policy context that may still affect the school nutrition
programme is the Department of Education’s planned changes to the definition of poor
learners across provinces. While such plans are not yet reality, we speculated that
implementation would affect poor and rich provinces in different ways. We argue that poor
provinces would have a sizeable number of poor learners in the poorest categories/quintiles,
thus reducing flexibility in extending this programme to poor secondary schools. In the case
of richer provinces, a relatively smaller number of poor learners would be located in the
poorest quintiles, thus increasing the probability of delivering to select poor secondary
schools. These speculations point to the difficulty of re-orienting the school nutrition policy
from a grade level focus to a comprehensive anti-poverty focus, irrespective of grade level.
The absence of discretionary funding bases in most provinces further encourages the view
that the school nutrition programme’s implementation would follow a forked logic, so evident
in all other funding polices at provincial education level.
In our review of the efficiency and effectiveness of spending, we identified a number of
factors that reduced the overall effectiveness of spending. Both poor funding prioritisation
and weak actual spending ratios contributed to the reduction of effectiveness in spending. We
argued against an interpretation that suggests that funding declines were the result of better
targeting or efficiency gains. Our review of the targeting practices of the provincial health
departments and provincial education departments brought to the fore the existence of dual
targeting of the same institutions. After 2000, both the school funding norms (education) and
provincial health targeting strategies (nutrition) aimed at identifying the poorest primary
schools. In such a scenario, the ideal would be a situation where a poor school that receives
affirmation through the school funding norms (by being placed in the poorest quintiles) would
also be on the school nutrition beneficiary list. Revisions to beneficiary databases by
provincial education departments in 2004 and 2005 suggest that these two targeting systems
may not have always identified the same schools. This represents a weakening of the impact
of anti-poverty programmes because of poor co-ordination across departments.
Evidence for the weak funding prioritisation of the school nutrition programme is also found
in the participating number of schools and learners. Given the constant presence of poverty
and the fact that the school nutrition programme never made space for other components of an
integrated and comprehensive nutrition strategy, declines in school and learner numbers
reflected the tight fiscal margins of this programme. During the period 1995 to 2003, the
number of participating schools was reduced by 15.4 per cent, while learner numbers were
reduced by 29.0 per cent. After 2000, the number of participating schools was increased from
16 200 in 2000/01 to 17 000 in 2003/04. This slow, but necessary increase after 2000, does
not match the extent of the reductions that took place in the 1995 to 2000 period. Similarly,
targeted learner numbers also experienced a decline from a “high” of 6.8 million in 1995/96
to 4.8 million in 2003/04. This represents a 29.0 per cent decline, and if we couple this to
variable targeting success, then the true impact of the school nutrition programme is further
shrunk. Learner numbers are rising because the data for 2004/05 show that 7.0 per cent more
learners actually participated in the school nutrition programme than originally planned. This
is directly related to the inefficiencies in targeting that we pointed out earlier and the directive
3
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
from the Department of Education that no participating school be removed from the
beneficiary lists.
What are the main challenges concerning the implementation of the school nutrition
programme? The re-orientation of the school nutrition programme from its historical grade
level focus (the old PSNP) towards an explicit affirmation as an anti-poverty measure is
arguably the greatest challenge. To achieve this requires clear policy and practical guidelines
on the identification of the main beneficiaries. This represents the intersection of the school
funding norms and the school nutrition programme and suggests that the future of both
programmes is intertwined. Therefore, in the absence of re-designed legislation in the area of
the school funding norms, efforts at developing a full-blown anti-poverty school nutrition
programme are frustrated. This explains why provincial education departments’ first policy
impulse was oriented towards a maximum extension of the nutrition programme to larger
number of primary schools. Finally, communities need to be a given a greater stake in the
delivery of healthy food. This not only enhances the budgetary value we all get from the
delivery of the school nutrition programme, but builds nutrition-conscious communities
whose actions stretch far beyond the programme confines of the school nutrition programme.
=====================================================================
SECTION 1: INTRODUCTION
Eight years after the birth of the school nutrition programme, and in the
context of spiralling food prices, the national Cabinet of South Africa chose
to affirm the school nutrition programme in the following manner (Cabinet
Statement, 25 July 2002):
It is in this context of a comprehensive approach to poverty eradication that Cabinet
examined the Report of the Committee of Inquiry into Comprehensive Social Security
and comments from the public. In addition to these issues, further work is being done
to examine the efficacy of increasing the age of child grant beneficiaries as well as
massive expansion and improvement in the efficiency of the school nutrition
programme.
Further Cabinet statements (in September 2002) re-affirmed Cabinet's
interpretation of the school nutrition programme as an essential anti-poverty
programme.iii Without further debate as to whether Cabinet's positions
define the school nutrition programme as a nutrition or anti-poverty
programme primarily, it is clear that the Cabinet statements seem to
support the further extension of the school nutrition programme.
In a different, but related context, researchers have estimated the impact of
social security on various measures of household welfare (Economic Policy
Research Institute, 2004). Table 1 shows the impact of a select number of
variables on household food shares and household basic food shares.
4
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Table 1: Household expenditure models of food shares (See full table in
the appendix)
Predictor variables
All food items
Coefficient P-value
Basic food items
Coefficient P-value
ln (household income per capita
-0.077619
0.000
-0.045139
0.000
Remittance received by household
0.000276
0.995
0.0114
0.674
0.08086
0.869
1.17706
1.25238
0.074
0.000
0.000
0.000
Years of education attained by household head 0.36863
0.000
State Old Age pension
1.52
0.000
Child Support grant
1.47498
0.002
Disability grant
2.49501
0.000
Source: Economic Policy Research Institute (2004: 78)
Restricting our attention to the receipt of social grants, EPRI (2004: 79)
notes that each thousand rand of annual state old age pension is associated
with an increase of 1.5 per percentage points in the share of household
spending on all food items. Similar and stronger patterns are found for the
child support grant and the disability grants. Although the impact of these
grants on household basic food shares is less, the overall positive pattern is
maintained. EPRI (2004: 75) summarises the evidence by saying
This is the most robust finding of the expenditure analysis-regardless of the type of
social grant, or how the food share is calculated, social grants are associated with an
increased allocation of spending in a manner that supports better nutrition (our
emphasis).
The link between social grants and better nutrition that is claimed by the
EPRI report is facilitated by the nature of the grant. Social grants are cash
transfers, which permit individual and household discretion over the use of
these grants. This does not imply that grant recipients always take decisions
that support better nutrition outcomes, but the space undeniably exists for
good decisions to be made. This contrasts sharply with the school nutrition
programme, which is an example of an in-kind transfer or benefit. Images of
schooling communities passively receiving school nutrition benefits are not
exaggerated. This is the case because the school nutrition programme has
not been transformed from a nationally imposed school feeding option, and
this makes effective linkages between the nutrition programme and strong
nutrition outcomes unlikely.iv Therefore, the link between school nutrition as
an in-kind transfer and better nutrition is not automatic, precisely because
of the absence of choice for beneficiaries.
Although Cabinet statements and academic commentators presently refer to
the school nutrition programme as an exclusively feeding programme, the
argument for the transformation of this programme that would enable
greater community participation remains strong. Community actors that are
directly linked with the school nutrition programme need nutrition
education programmes and advocacy. Communities also need ownership of
these programmes. We motivate this because we believe that strong links
between the school nutrition programme and good nutritional (and health)
outcomes would support and enhance the reality of budgetary increases to
5
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
the school nutrition programme allocations. The transformation of the
school nutrition programme would therefore coincide with attempts at
reforming the efficiency and effectiveness of spending on this vital
programme.
Road map of the occasional paper
This paper is structured in the following way. Section 2 reviews the local
literature on the implementation of the school nutrition programme. It also
discusses the implications of the local literature and how this impacts on the
development of specific objectives for the present paper. Section 3 discusses
three different public finance contexts within which the school nutrition
programme operated. The first sub-section deals with the inception of the
nutrition programme and how it was affected by the prevailing fiscal
circumstances and policies so powerfully represented in the White Paper on
Reconstruction and Development 1994. The second sub-section examines the
school nutrition programme under the integrated nutrition programme and
compares trends in real growth of consolidated provincial health spending to
real growth of the school nutrition allocation. Reference is also made to
conditional grant reforms that were introduced in this period. The final subsection considers the school nutrition programme in its present context
where it operates as a stand-alone conditional grant administered by the
Department of Education. Section 4 discusses service delivery and
implementation issues. We examine issues pertaining to the efficiency and
effectiveness of spending through available time-series data. We also
undertake an analysis of differences in targeting methodologies between
provincial health and provincial education departments and how these have
affected the overall effectiveness of spending. Section 5 offers concluding
remarks.
SECTION 2: REVIEWING THE LOCAL LITERATURE ON THE
IMPLEMENTATION OF THE SCHOOL NUTRITION PROGRAMME
The school nutrition programme (formerly known as the primary school
nutrition programme-PSNP) was introduced in September 1994 as one of the
Presidential Lead Projects. In its original guise, the White Paper on
Reconstruction and Development 1994 (page 46) described the aims of the
PSNP as follows:
To contribute to the improvement of education quality by enhancing primary pupils'
learning capacity, school attendance and punctuality and contribute to general health
development by alleviating hunger. Educating pupils on nutrition and also improving
nutritional status through micro-nutrition supplementation. Parasite eradication where
indicated. To develop the nutrition component of the general education curriculum.
It is not difficult to understand why there was so much debate and
discussion about the status of the school nutrition programme because both
feeding and broader nutritional health outcomes were linked. To help us
better understand the nature of these debates, we now turn to a critical body
of knowledge, which had developed as government-sponsored reviews and
independent academic analyses of the school nutrition programme.
Although this section does not present an exhaustive account of such
research, we aim to develop the main ideas and research findings from the
6
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
established local body of work on the school nutrition programme.v
Researchers' views and findings are discussed with a view to




Their overall perspective on the school nutrition programme.
The stated benefits and costs of the school nutrition programme.
The conditions that would enhance the successful implementation of
the school nutrition programme.
The future of the school nutrition programme.
The Child Health Unit (1997) regarded the primary school nutrition
programme (and particular its manifestation as a feeding programme) as
only one component of a comprehensive nutrition strategy. The latter would
include school feeding, the mass application of de-worming medication, the
delivery of micronutrients, control of parasitic worm infections, family
planning, life skills, and education aimed at reducing tobacco and alcohol
use. Factors that counted in the favour of school-based nutrition
programmes were access to ready-made infrastructure and the fact that
there are normally more teachers than nurses, as well as a greater number
of schools than clinics. From a cost perspective, the researchers argued
compellingly that an exclusive focus on school feeding is counter-productive,
too costly, logistically difficult to implement, and that school feeding was
locking in scarce human and material resources. An unconvincing body of
research on the potential benefits of school feeding further fuelled a scathing
review of the stand-alone school feeding programme. The Child Health Unit
(1997) believed that other components of a comprehensive strategy should
be promoted, the school nutrition programme needed to be extended to the
pre-school cohorts, and the active promotion of community-based nutrition
programmes to enhance education and awareness of nutrition matters.
Finally, these researchers argued for the down-scaling of school feeding
through narrower targeting practices, while up-scaling neglected areas of an
integrated school nutrition response. Narrow targeting would contain the
cost of school feeding, while providing financial scope for neglected areas of
an integrated school nutrition response.
The Louw, Bekker, and Wentzel-Viljoen 2001 evaluation (hereafter referred
to as Louw et al.) focused on assessing the effectiveness and efficiency of
school feeding in terms of specific implementation and operational issues.
The perspective of this group of researchers was that if the school feeding
programme could be properly designed and effectively implemented, the
benefits of school feeding far outweigh the investment made by government
(Louw et al., 2001: viii). To support their view of school nutrition, Louw et al.
(2001: 156 and 157) quoted international studies that record the positive
impact of school feeding on school attendance, community involvement in
schooling, and reduced malnutrition rates.vi Results from their sample study
also indicated gains in areas of concentration spans, punctuality, school
performance, and the alleviation of temporary hunger. Although no explicit
policy costs were mentioned, the researchers believed that service level
agreements between education and health, standardisation of menu and
costs of menu, a simplified programme implementation, and the monitoring
of processes would create favourable conditions for implementation. While
speculatively posing the question about the future of the school nutrition
programme, Louw et al. (2001: 241) recommended that the then primary
7
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
school nutrition programme continues as a nutrition programme as opposed
to being a social relief or anti-poverty programme.
Steyn and Labadarios (2002) reviewed the implementation of nutrition policy
by analysing components of the Department of Health's integrated nutrition
programme (INP). The researchers' perspective on the school nutrition
programme was that as long as unemployment remained high,
"supplementary school meals for needy school children should be continued"
(page 333). Although their paper did not deliberate on the policy costs of the
nutrition programme, it bemoaned the lack of progress that had been made
since the first full-scale review (the Child Health Unit 1997 report) was
conducted in 1996. They also argued that problems relating to targeting and
learner/school coverage had not been solved. In terms of the future of the
nutrition policy, the researchers argued that cost-effectiveness analyses
should be conducted to determine physical and intellectual performance
benefits.
Following all these reviews, the Department of Health's (Kloka, 2003: 1)
response was very instructive. It argued that
"The PSNP was primarily designed to provide direct services to primary school
learners to reduce hunger and to alleviate the effect of malnutrition on their learning
capacity and not to improve the nutritional status of school learners "
(Department of Health's own emphasis, page 1).
Furthermore, it cited the research done by Louw et al. (2001) indicating that
school feeding contributed to household food security and that it was not
seen by educators as an infringement on learning time, or as requiring too
much time. The Department of Health also endorsed many of the
recommendations from the Louw et al. (2001) report and concluded that
standardised menu options, refined poverty targeting, implementation
simplicity (feed early morning), a food safety monitoring system, and a
national school feeding data base would aid the implementation of school
nutrition programmes. The future of the school feeding programme was
mapped out in 2003 as involving a transfer of responsibilities to the
education sector. The reasons given were the education outcomes of school
feeding; the fact that school feeding was implemented in schools; and to
facilitate the inclusion of school feeding into the broader context of
education development (Department of Health, 2003: 2).
Hunter, May, and Padayachee (2003-hereafter referred to as Hunter et al.)
examined the contested issue of poverty reduction strategy papers (PRSPs) in
developing countries and attempted to outline South Africa's "poverty
reduction" strategy. Their perspective on the nutrition programme was that
services, which are preventative in nature (such as sanitation and nutrition),
have benefits that outweigh their costs and should be expanded. They did
however point to serious implementation problems concerning poor learner
coverage, the lack of good data, the absence of a national nutritional
surveillance system, leakage of funding, and ineffective monitoring and
evaluation of nutrition programmes. Based on their understanding of
poverty reduction processes, they argued for improvement in implementation
on two grounds. Targeting processes needed to be refined to permit
maximum coverage of the needy, and pro-poor targeting policies should be
8
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
better co-ordinated so that the benefits of the nutrition programme are
supplemented and enhanced.
Brand (2004) discussed the primary school nutrition programme within the
context of national food security generally, and household food security more
specifically. His perspective on the nutrition programme was that it should
not be regarded as government's premier nutritional programme in respect
of children (Page 114). According to Brand (2004), the services of the
programme were too intermittent and it covered too small a part of the year
to qualify comprehensively as a nutritional programme. In the absence of
other more comprehensive programmes that would realise children's right to
nutrition, Brand recognised the primary school nutrition programme as the
"only programme that is explicitly (even if only partially) intended to directly
advance children's right to basic nutrition." However, if government is intent
on regarding the school nutrition programme as the premier programme
realising children's right to food, then the policy would impose a cost that is
not worth bearing according to Brand. Brand argued that the absence of a
comprehensive social security net weakens the impact of the school
nutrition programme. In addition, the fact that it targets only a small portion
of needy children further weakens its claims as a comprehensive nutrition
programme. He argued for the continuation of narrow targeting and insisted
that instead of increasing the number of beneficiaries, the emphasis should
be placed on improving delivery to the present beneficiaries.
The broader implications of research: developing objectives for
the present paper
Two issues stand out from the review of the local literature, namely the
classificatory status of the nutrition programme and the question (or
implications) of the policy costs of the nutrition programme.
With regards to the first question, researchers either regard the school
nutrition programme as a feeding scheme or alternatively as a more
comprehensive nutrition programme. Consistent with the idea of calling the
school nutrition programme a nutrition programme would be the (cost and
targeting) containment of the school feeding component and the
corresponding up-scaling of other aspects of a comprehensive school
nutrition strategy. This thought does not only speak to nutritional
considerations, but goes to the heart of efficient and effective spending of
scarce State resources. It suggests a process where school feeding targeting
would be stricter and fewer schools and learners would be included,
especially given explicit budgetary constraints. This would create the
financial space for other (perhaps more vital) elements of an integrated
school nutrition response. The obverse of this argument is also compellingly
simple: if we regard the present school nutrition programme primarily as a
feeding programme, then the arguments for tighter targeting and
containment of the overall costs of school feeding are considerably
weakened. In the absence of an appropriate balance between feeding and
other nutritional considerations, school feeding would operate in the same
financial and fiscal space as funding policies that aim to reach the broadest
majority of poor learners. There would be very little justification for servicing
only a small percentage of poor learners, given overall levels of poverty in
South Africa.
9
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Disqualifying the school nutrition programme as a nutrition programme
would therefore strengthen the argument for the extension of the programme
to more poor learners in primary schools, and eventually to the extension of
this programme into poor secondary schools. Most researchers in the South
African context would not object to calling the school nutrition programme a
school feeding scheme and therefore the above comments must apply. Thus,
one of the tasks of this paper is to carefully follow output and service
delivery trends of the school nutrition programme and to assess whether
extension to more primary learners and poor secondary schools is feasible.
This must obviously be studied within the context of the Department of
Education's most recent pronouncements on the distribution of socioeconomic groupings in schools.vii
The second question concerns the perceived "policy costs" of the school
nutrition policy. Most researchers in the South African context argue that
the broader policy benefits of the school nutrition programme outweigh its
putative costs. These assertions are made regardless of the fact that no costbenefit analysis or cost-effectiveness studies have been done. The Child
Health Unit (1997) study is the only review that attempted to understand
how the policy costs reduced the effectiveness of the school nutrition
programme. Supporting the overall positive assessment of the school
nutrition programme is the reality of large real budgetary increases to the
school nutrition programme, which suggest a medium-term to long-term
future for the nutrition programme.viii Problematic as these developments
may be to some, the continuation of the nutrition programme warrants
attention to service delivery and implementation issues. We are particularly
interested in understanding how well the school nutrition policy fared in
terms of the efficiency and effectiveness of spending.
A note on the research methods
Unlike the systematic sample studies of Louw et al. (2001), information on
service delivery developments was exclusively obtained from eleven
government agencies, namely the national Department of Education, the
national Department of Health, and the nine provincial education
departments (PEDs). Because much of our focus was on the most recent
situation and given the fact the historical data on the school nutrition
programme were available, we did not interview officials from the provincial
departments of health. The service delivery information is produced by
government, which means we have relied on the integrity of the respective
departments concerning the accuracy and factual correctness of the data.
We were unable to verify whether this information is correct as this would be
beyond the scope of our available resources. Where data appeared
"unusual", especially in a time-series context, we have followed this up with
the respective departments. We restricted service development information to
the following areas for the 2004 and 2005 academic years:



Discretionary funding of the school nutrition programme by PEDs.
Targeting procedures followed by PEDs and how these differ from
previous targeting practices under the provincial departments of
health.
Frequency of feeding and total number of feeding days per calendar
year.
10
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA



Budgetary and output trend information concerning the school
nutrition programme from 1994/95 to 2005/06.
Estimates of total cost per learner and administrative cost per learner
in 2004/05 and 2005/06.
Basic monitoring and evaluation questions.
Apart from the desktop study of the main review studies that were
conducted about the school nutrition programme, we also did a
comprehensive search of relevant newspaper coverage since 1994. A list of
newspapers that were used is provided in the reference list.
The budget trend analyses use 2000/01 and 2004/05 as base years. Table 2
provides the CPIX deflators based on these respective base years.
Table 2: CPIX deflators using 2000/01 and 2004/05 as base years
Financial Year
2000
2004
1995
0.691867
0.537704
1996
0.74029795
0.5753431
1997
0.809485
0.629114
1998
0.867768
0.67441
1999
0.927644
0.720944
2000
1
0.777178
2001
1.066
0.828471
2002
1.170468
0.909662
2003
1.234844
0.959693
2004
1.286707
1
2005
1.340749
1.042
2006
1.411809
1.097226
2007
1.486634
1.155379
Source: Personal communication with National Treasury, 2003 (authors' own
calculations)
Note: The annual index for 1996 refers to a CPI figure, because government officially started
using CPIX in 1997.
SECTION 3: THE CHANGING PUBLIC FINANCE CONTEXTS OF
THE NATIONAL SCHOOL NUTRITION PROGRAMME
Although public finance contexts do not correspond to a neat linearization
scheme, it is undoubtedly the case that fiscal policy and budget policy in
South Africa were subject to fundamental changes over the eleven year
period considered in this paper. In some instances, changes were marked by
the abandonment of systems, while in other instances, changes in spending
patterns dictated the definition and establishment of new contexts.
The School Nutrition Programme and the White Paper on
Reconstruction and Development and Zero-Based Budgeting (ZBB)
The macro-economic impetus of the White Paper on Reconstruction and
Development 1994 (hereafter referred to as the RDP White Paper) can partly
be traced to the policy-making process of the previous regime. The relevant
policy context was the previous regime's struggle with run-away
consumption expenditure and macro-economic aggregates that pointed
11
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
strongly to the "overbearing" role of the State in the economy. The policy that
was developed to address these issues was the White Paper on Privatisation
and Deregulation in the Republic of South Africa (1987). The conceptual gist of
this White Paper was to lessen significantly the participation of the State in
the private economy and to create more space for private sector
participation. In 1985, the public sector's contribution to the Gross
Domestic Product was 38.1 per cent, of which central government
contributed 26 per cent (White Paper on Privatisation and Deregulation in the
Republic of South Africa, 1987: 4). Similar "unfavourable" aggregates were
quoted such as government's domination in the use of private investments,
which the White Paper argued, crowded out the private sector's potential
investment base.
Five years after the launch of the White Paper on Privatisation and
Deregulation in the Republic of South Africa, many of the key targets of the
policy had not materialised. The current expenditure to GDP grew from 17.4
per cent in 1980 to 28.9 per cent in 1993/94 (South African Reserve Bank,
1996). Capital spending declined in the corresponding period, while
employment in the private sector decreased by 11.7 per cent in the period
1989 to 1994 (SARB, 1996). One of the stated aims of the policy was to move
factors of production to the more productive private sector to encourage
allocative efficiency. The net result of this action was a major public
investment slowdown and some economists argued that falling parastatal
investment contributed to the recession that lasted from 1989 to 1993
(Makgetla, 1995:69).
Given the failure of the apartheid regime in controlling government
consumption expenditure, it fell to the post-1994 government to confront
this problem. The RDP White Paper states explicitly that government's
immediate challenge was to fund and staff the RDP without exacerbating
already high levels of government debt (RDP White Paper, 1994: 21):
Increasingly, the market evaluation of such a situation [of high government debt and
high consumption expenditure] was that the government could not curb expenditure,
dissaving would continue, the balance of payments would be adversely affected and
inflation would rise. As a result, interest rates rose and increased the government debt
burden. In the context of macro-economic instability, other crucial objectives can be
undermined (Our insertion).
Government's commitment to reducing consumption expenditure included
the attempted re-direction of expenditure from consumption expenditure to
capital expenditure, not filling public service vacancies and forward planning
on all projects and programmes. Re-prioritisation of expenditure was
supposed to have been facilitated by the introduction of the RDP Programme
Fund.ix Resources for the RDP Fund were made up of international donor
contributions and "top-slicing" from departmental budgets. Departments
could in theory access these resources, but had to demonstrate commitment
to expenditure re-prioritisation in line with government's key objectives. The
RDP Programme Fund was therefore conceptualised as a mechanism that
would start and direct a process of re-prioritisation towards government's
vital new priorities.
Segal (1996) noted that the primary users of the RDP Fund were free health
care, the primary school nutrition programme, education and constitutional
12
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
development. Furthermore, the RDP Fund was intended as a wedge to get
departments to re-assess their core business and priorities. The RDP Office
was closed in 1996 amidst concerns about effective service delivery
(Business Day, 1997). The two key issues that had been mentioned as the
reasons for the closure of the RDP Office were the complicated relationship
between the RDP Office and national cabinet ministers, as well as the lack of
expenditure re-prioritisation through the RDP Fund. In relation to the latter,
it was generally accepted that departments, by applying and using RDP
Fund monies, were in essence postponing expenditure re-prioritisation. A
good example of this type of behaviour was evident in the pronouncements
of the then Minister of Health who argued that cuts in provincial health
budgets would be made good by the availability of more funds from the RDP
Fund (Sidley, 1995). In addition, the health department would have asked
the department of finance for more resources to cover the gaps left by the reallocation of resources. The closure of the RDP Office was therefore seen as
an admission that the RDP Fund contributed to the lack of expenditure reprioritisation (Business Day, 1997). Projects that were funded under the
banner of the RDP Fund were now integrated into the relevant departmental
budgets and would have been funded since the 1997/98 financial year from
own departmental funds.
The zero-base budgeting system (ZBB) was the budget al.location system
that was favoured by government. This system requires that spending
agencies at the national and provincial level annually review their spending
items and justify their present and future funding based on approved
national priorities. Abedian et al. (1997: 91) argued that spending agencies
were required to define their policies, quantify their target outputs, model
the costs of their programmes, and submit these as business plans to
various treasuries.
The ZBB system was regarded as a natural ally to the RDP Fund process of
re-prioritising expenditure (RDP White Paper, 1994: 17):
There will be a tendency for inertia when existing programmes of the government at
all levels are reviewed for the purpose of re-directing expenditure and resources. In the
planning and budgeting process, it is therefore essential that departments and tiers of
government place all programmes on an equal footing in allocating funds, staff and
resources. Programmes should not be preferentially funded and staffed purely
because they have been previously established (Our italics).
The commitment to the ZBB system did not betray a short-sighted focus on
the annual budget, but government believed that it was possible to link the
ZBB system to multi-year budgeting. At that time, South Africa, like most
other countries, budgeted on an annual basis. The RDP White Paper reflects
on this link in the following manner (page 30):
At present the Budget is drawn upon an incremental basis. Allocations to the different national
line function departments are decided on the basis of what the allocations were in the previous
year. The government will introduce a zero-base budgetary process, by which national line
function departments, provinces and other institutions of government will motivate their
programmes, and on this basis determine their budgetary requirements. In addition, the
government will introduce a multi-year budgeting process, by which budgets are drawn up for a
period of three years on the basis of ongoing programmes.
13
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Abedian et al. (1997: 86) noted that although the ZBB system was
government's preferred budget allocation system, setting the budget to zero
each year is not practical as some programmes run over more than one year.
ZBB was thus a useful call for reconsideration of all expenditure items, but
was never implemented.
What was the impact of these contexts on the implementation of the school
nutrition programme? The impact of these contexts was twofold: a strict
insistence upon non-expansion of staff levels for designated RDP Projects,
and budget allocations that did not keep pace with inflation. Newly formed
provincial governments were therefore required to implement a project that
would require specialist staff, but for which they did not receive any
additional personnel funding. The Child Health Unit (1997) bemoaned this
strategy and argued that the result was an inefficient use of human
resources because school feeding consumed much of the available scarce
human resources. The strategy of availing earmarked funds to a subnational sphere to further national objectives had not taken on board the
human resource requirements for successful implementation.x It would
appear that the roots of this conception can be found in the immediate post1994 fiscal context where government deliberately tried to contain
government consumption expenditure.
The second recognisable impact of this period was the real reduction in the
school nutrition budget over the first few years. For the period 1995/96 to
1997/98, the school nutrition allocation declined by approximately 8.0 per
cent in real terms. Over the period 1995/96 to 2000/01, the school nutrition
allocation declined in real terms by 4.2 per cent. We are aware that great
caution is required in interpreting the large real declines purely from the
point of view of government's fight against consumption expenditure. This is
the case because in 1996, government's fiscal and budget policy were made
with reference to its macro-economic policy, Gear. However, the thrust of
Gear was also to moderate spending claims on the national purse and thus
the overall effect would have been the same. If we interpret non-expansion of
staff levels and real declines in the school nutrition allocations together, the
containment of the funding of the school nutrition programme was achieved.
With the closure of the RDP Office in 1996 and budget reform focusing
strongly on the development of a multi-year budget framework, the ZBB
system was abandoned in favour of the medium term expenditure framework
(MTEF). With government accepting the medium term nature of expenditure,
ZBB would no longer have sufficed as a budget allocation system.
Furthermore, government's wishes to give more predictability and stability to
the budget system would have made the MTEF the preferred candidate. So
by the start of the 1998/99 financial year, two crucial components of the
post-1994 fiscal set-up, namely the RDP Fund mechanism and the ZBB
system, were abandoned.
14
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
School Nutrition Programme under the Banner of the Integrated
Nutrition Programme Conditional Grant
In the 1997 White Paper for the Transformation of the Health Sector, the
Department of Health declared four components of an integrated nutrition
strategy. These were the health facility-based component, a communitybased component, a nutrition-promotion programme, and a national
nutrition surveillance system. The school nutrition policy was categorised
under the community-based component. The 1998 health document
Integrated Nutrition Policy: a Foundation for Life directed that nutrition policy
should be targeted towards vulnerable children and women. In addition, it
argued for integrated approaches and a break with fragmented food-based
approaches.
The integration of the primary school nutrition programme into the
integrated nutrition programme was a direct consequence of the 1997 White
Paper for the Transformation of the Health Sector. This integration did not
change the status of the primary school nutrition component as being a
direct in-kind transfer and therefore all the comments about the limitations
of such funding still apply. However, we feel that there are two other issues
that now require closer scrutiny. The school nutrition programme's
integration under the integrated nutrition programme happened over a sixyear period, namely 1998/99 to 2003/04. Two distinct periods in the
evolution of public finances in South Africa are contained during this period
and so the first issue relates to how the school nutrition programme
navigated these contexts. We answer this question by using consolidated
provincial health budgets as anchors in our interrogation of broader
spending patterns. The second big issue concerns the development of
conditional grants and the initial problems that were experienced with this
funding mechanism. Towards the end of the section, we briefly consider how
both these contexts affected prioritisation and implementation of the school
nutrition programme.
The discussion of the school nutrition programme under the banner of the
integrated nutrition programme does not mean that the period 1998/99 to
2003/04 was undivided in a public finance sense. In fact, social services
spending struggled under the grip of Gear in the period 1996-2000, although
the extent of reductions was not uniform across government departments. xi
There is an overall consensus today that after 2000, government was in a
position to accelerate spending in areas that were worst hit by the fiscal
austerity measures during the Gear period.
Figure 1 compares the real year-on-year growth rates of consolidated
provincial health budgets, integrated nutrition grant spending, and primary
school nutrition spending over the period 1998/1999 to 2002/03.
15
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
6.0%
0.0%
4.0%
-2.0%
2.0%
-4.0%
0.0%
-6.0%
-2.0%
1999
2000
2001
2002
4-Year average -8.0%
-4.0%
-10.0%
-6.0%
-12.0%
-8.0%
-14.0%
Real year-on-year change
for PSNP
Real year-on-year change
for health and INP
Figure 1: Real year-on-year change in the budget allocations of
consolidated provincial health, integrated nutrition grant and primary
school nutrition programme, 1998/99 to 2002/03 (2000 Rands)
Health
INP
PSNP
Financial Years
Source: Personal communication with Department of Health 2005 and National
Treasury's Intergovernmental Fiscal Reviews 2001, 2003 and 2005
While we have alluded to the fact that the integrated nutrition grant period
includes possibly two different moments in the evolution of fiscal
management in South Africa, this appears to apply exclusively to
consolidated provincial health spending. We see clearly that in 1999,
provincial health budgets were still in real decline, but appeared to pick up
steam in the post-1999 period. Thus, for the period defined in the chart
above, provincial health spending recovered strongly and registered a fouryear average annual growth of approximately 1.4 per cent. Blecher and
Thomas (2003: 275) noted that although real increases took place in the
post-2000 period, most of these increases appeared to have been absorbed
by rising wage costs and medical inflation. Such real increases did not
therefore result in improved service coverage and quality.
These comments are particularly useful in the context of financing trends for
the integrated nutrition grant and the primary school nutrition programme.
The real year-on-year trend for the primary school nutrition programme
conforms more to general developments in social expenditure over the 19962000 period. Over this period, the school nutrition programme showed
consistent negative declines but unlike overall health spending, there was no
notable growth after 2000. In fact, over the period depicted in the chart
above, the school nutrition programme sustained real average annual losses
of 7.0 per cent. Although the integrated nutrition grant suffered similar
losses over the same period (4.4% four-year decline), the primary school
nutrition programme bore heavier losses. However, the overall trend
suggests that grant funding for nutrition-related interventions suffered
badly, even in a context where overall health funding was beginning to
increase.
We examine longer term trends of the school nutrition programme in later
sections, but we can observe that the real decline in spending could not be
blamed solely on poor spending performance. Although the spending rate on
the school nutrition programme was highly variable and there were clear
inefficiencies in spending, collectively over the 1998/99-2002/03 period,
provinces managed to spend on average 86.0 per cent of budgeted
16
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
allocations. These represented levels of spending that were much higher
than comparable spending on conditional grants in provincial education
departments. In the 2000 Budget Review, the National Treasury recognised
the problems of conditional grant spending and discussed planned changes
to the conditional grant framework (page 157):
Over the past two years, significant underspending of certain conditional grants
resulted in rollovers. These rollovers reflect not only a lack of experience with
administering such grants but also insufficient capacity at both the national and
provincial levels to monitor and implement these programmes. Within national
departments, conditional grants have been insufficiently integrated with strategic
planning and budget processes. The roles and responsibilities of national departments
and grant recipients have often been unclear, further complicating grant management.
These
points
were
further
discussed
in
National
Treasury's
Intergovernmental Fiscal Review 2001 (113) where confusion over
accountability, poor design and planning, inadequate transparency in
allocations, too many conditional grants, and poor monitoring were pointed
out. It is our view that these problems were a symptom of the state of
intergovernmental fiscal management at the time. Barberton (2002a: 2)
argued that the period 1996-2000 was a particularly challenging time as
national and provincial departments were learning how to manage
conditional grants more effectively. He further alleged that national
departments were slow in assisting provinces in key areas of fiscal
management, thus delaying effective implementation of grant funding.
Although new cross-sphere budget institutions were introduced such as the
Budget Council and various MINMECsxii, these institutional gains required a
longer time horizon before actual improvements in inter-governmental fiscal
management could be observed.
Barberton (2002b: 213) noted that at the start of the 2001/02 financial year,
government increased the transparency of the Division of Revenue Act by
including detailed conditions for the management and use of conditional
grants.xiii Furthermore, by 2002, all the components of the expanded
framework for planning, reporting, and review that were developed in the
Public Finance Management Act (PFMA) were in place. Finally, at the start of
the 2003/04 financial year, the entire budget process would have been
properly sequenced, indicating the significant progress that was made in
improving inter-governmental fiscal management. Proper sequencing implies
that the entire budget process is anchored in the definition of political goals
and that the strategic planning process, the actual budget and
implementation, review and reporting follow accordingly. Budget
implementation would require monthly actual spending reports as well as
quarterly service delivery reports. Such tools should enable careful
monitoring of conditional grant spending and a co-operative governance
approach to solving problems if these are prominent. Control over spendingif not necessarily service delivery dynamics-had apparently been achieved at
the tail end of a long budget reform process.
Variable spending patterns in the school nutrition programme appear to
vanish at the end of 2001/02. Since the start of the 2002/02 financial year,
grant spending on the school nutrition programme remained consistently
high. Although the passage of time offers an alternative explanation for the
stability of spending since 2002/03, we cannot rule out the possibility that
17
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
reforms to the conditional grant framework had a definite impact on the
most recent spending patterns. We review output trends in the next main
section, but we can add that stable learner and school numbers would also
have contributed to the stable spending that took place since 2002/03. So
while the school nutritional allocations continued to decrease in real terms
in successive budget years, stable and consistently high spending rates
since 2002 partially began to compensate for the severe funding neglect of
the programme since its inception.
The Department of Education's inheritance of the school nutrition policy was
therefore marked by two features: budgeted school nutrition allocations that
consistently suffered real declines and a conditional grant framework that
was now much more enabling in promoting effective and efficient
implementation.
School Nutrition Programme as a Stand-alone Conditional Grant
with the Department of Education
In discussing the transfer of the primary school nutrition programme from
the Department of Health to the Department of Education, the national
Department of Health (following a Cabinet decision) pointed to



The education outcomes of school feeding;
The fact that school feeding is implemented in schools, which is the
functional responsibility of Education; and
Facilitation of the inclusion of school feeding into the broader context
of education development (Kloka, 2003: 2).
Our task in this section is to sketch the prevailing funding and policy
contexts in provincial education, and to understand how these may impact
on the implementation of the school nutrition programme. Based on this
review, we speculate about the funding future of school nutrition as well as
the implications of the extension of the national school nutrition programme
(as it is now called) to secondary schools serving poor learners.
How did consolidated provincial education funding fare over the period
1995/96 to 2005/06? Figure 2 details the real growth of consolidated
provincial education spending over this period.
18
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Rand (million)
Figure 2: Real growth of consolidated provincial education expenditure,
1995/96 to 2007/08 (2004 Rands)
80000
70000
60000
50000
40000
30000
20000
10000
0
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Financial Years
Source: Budget Review
Expenditure 2005/06
2000/IGFR
1995-2003/Provincial
Estimates
of
Note: The CPIX value for 1996 is a calendar year value (7%) taken from Budget Review 2000
(page 40)
There are three features that defined consolidated provincial education
funding over this period. Firstly, the upswing in expenditure in 1996 was
caused by the historic wage settlements between public service unions and
government. The brevity of this upswing reflects the "accidental" and
unplanned nature of these increases. Secondly, the period between 1996
and 2000 was characterised by real declines in provincial education
spending. The post-2000 period was characterised by a moderate, but steady
increase in provincial education funding. The length of the post-2000 curve
suggests that these changes were intended, unlike the experiences in 1996.
Wildeman (2003) argued that post-2000 expenditure growth was defined to
accommodate increases in non-personnel and capital funding. He also
argued that such funding changes were conceptualised within the ambit of
narrow targeting. Such targeting had the effect of containing the cost of
post-2000 redress and allowed provincial education departments to stay
within their spending envelopes.
An abiding feature of post-2000 education funding has been the focus on
programmes and activities that -when compared to the overall funding base
in provincial education-are relatively small. In 2005/06, personnel
expenditure still consumed on average approximately 90 per cent of
provincial spending, while capital accounted for 4.1 per cent. Non-personnel
and non-capital expenditure would have made up the rest and yet this is the
base from which most of the post-2000 policies is funded. The first
implication of the school nutrition's transfer to education is that the
programme would become part of a context where real increases have been
made available to relatively small programmes. Under the supervision of
education, the funding fortunes of the national school nutrition programme
are therefore set to change from negative to positive real gains. Indeed, if one
examines the growth of the national school nutrition programme over the
medium term, it is projected to grow at a real average annual rate of 6.4 per
cent. This is a far cry from the drips it was accustomed to under the RDP
Fund and the integrated nutrition grant mechanisms.
Further to the context within which the national school nutrition programme
is inserted would be the barrage of changes related to school funding. The
19
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
school funding norms, which were first implemented in 2000, were intended
to guide the allocation of non-personnel/non-capital spending within
provincial education departments. Inequitable funding bases meant that
some provinces appeared to have implemented this policy more successfully
than others (See Wildeman, 2000, 2001, 2003 and 2004). One of the biggest
concerns was the unequal treatment of poor learners across provincial
departments. Political pressure and pressure from civil society groupings
resulted in a fundamental review of the cost of schooling, including a
sustained focus on the school funding norms (Department of Education,
2003a and 2003b). The net result of all these policy debates was the positing
of "adequacy" amounts for school funding allocations and the declaration of
school-fee free schools (Department of Education 2004). Wildeman (2004)
argued that these proposed changes did not herald the beginning of
adequacy funding, but the Department of Education was merely using its
moral clout to define what it considers as decent funding allocations.
What is different, following the suggested policy changes, is the methodology
of determining the definition of poor learners across provincial boundaries.
Using national income and expenditure data, the Department of Education
proposed a division of learners into five quintiles that follows the national
distribution of income categories or quintiles. In poor provinces, this has the
effect of increasing the number of poor learners that need to be considered
for redress. In rich provinces, this has the impact of reducing the numbers
of learners in the poorest categories, and results in large number of
previously defined poor learners being located in richer income brackets.
Table 3 demonstrates this situation and reflects earlier work done by the
Department of Education in fitting provincial income distributions within
nationally-defined income brackets.
Table 3: The distribution of national income quintiles at provincial level
according to Statistics South Africa 2000 data (%)
National quintiles
Eastern Cape
Free State
Gauteng
1 (poorest)
34
33
7
2
26
20
11
3
18
16
18
4
10
14
28
5
(least
poor)
11
18
35
Total
100
100
100
KwaZulu Natal
Limpopo
Mpumalanga
19
27
14
22
25
23
22
22
25
21
15
21
16
10
17
100
100
100
Northern Cape
North West
18
20
17
19
21
23
20
23
23
15
100
100
29
20
40
20
100
100
Western Cape
4
10
16
South Africa
20
20
20
Source: Department of Education, 2003b: 17
The proposed changes do not only impact on the implementation of the
school funding norms, but affect every policy that targets poor learners in
provincial public schools. Changes to the school funding norms do not
20
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
assume a distinction between poor primary and secondary school learners.
Although the South African Schools Act of 1996 is relatively clear in terms of
its definition of "compulsory education", the school funding norms in
targeting recurrent non-personnel/non-capital funding do not draw a
compulsory/non-compulsory distinction.xiv At a minimum, this would
suggest that reference to the "primary school nutrition programme" would be
problematic because the re-distribution of learners in table 3 does not
distinguish between primary and secondary school learners. Thus, the
primary axis of the re-definition of poor learners is not the grade level of
learners, but broader poverty conditions that define the circumstances of
learners and their schools. The net effect of such changes would require the
re-definition of the school nutrition programme away from grade level
(primary school nutrition programme) to a school feeding programme that
serves poor primary, secondary, and Grade R learners. This is consonant
with the call by the African National Congress for the extension of the
programme into poor secondary schools (Martin, 2002).
This theoretical re-definition of the scope of the school nutrition programme
is unlikely to be implemented without significant new tensions emerging.
Chief among these would be the net funding implications of the redistribution of poor learners within provincial education departments. Table
3 shows that poorer provinces have more learners who qualify for immediate
redress funding, and therefore the cost of maintaining the nutrition
programme would be higher. These provinces have never quite managed to
improve their personnel/non-personnel ratios, and therefore the scope for
radical expansion of the school nutrition programme is limited. Under such
circumstances, the question must be asked whether extension of the school
nutrition programme should not rather focus on bringing more primary
schools into the scheme. This would sacrifice poor secondary schools' claims
to participation in the scheme and water down the intended aims of pro-poor
learner funding.
In the case of the traditionally affluent provinces, comparatively smaller
numbers of poor learners are located in the poorest income categories. The
financial pressure to deliver to these groups of learners would be relatively
less for richer provinces. This opens the possibility of defining the prime
beneficiaries of the school nutrition programme across schooling phases and
conforming to the poverty definition of participating schools. If this logic is
correct, the school nutrition programme in richer provinces would be able to
accommodate both poor primary and secondary schools. All of this suggests
that the school nutrition programme is projected to grow strongly over the
medium term, but that the source of the growth would differ across
provinces. Poor provinces are likely to extend the school nutrition
programme to more poor primary schools, while traditionally rich provinces
would appear to be in a better position to accommodate both schooling
phases.
This does not negate additional funding pressures: in poor provinces,
nothing prevents agitation for the extension of the programme to poor
secondary schools. Neither does the actual extension of the programme to
secondary schools in rich provinces eliminate funding problems with regards
to excluded poor primary schools. But it does appear that well-intended
21
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
policies and reforms will likely lead to further fragmentation of national
policy ideals.
The next section considers trend data that speak to the efficiency and
effectiveness of the school nutrition programme.
SECTION 3: SERVICE DELIVERY AND IMPLEMENTATION ISSUES
This section deals with service delivery and implementation trends.
Conceptually, the first sub-section discusses the efficiency of spending by
examining the deviation between budgeted and actual expenditure. We also
re-visit interpretations about the size and magnitude of the school nutrition
budget. We continue our probing of efficiency by comparing overall per
learner costs across provincial education departments. The second subsection assesses the effectiveness of spending. We start by discussing
changes in learner and school targeting and then shift attention to learner
and school output data. The third sub-section offers summary remarks
about the efficiency and effectiveness of the school nutrition programme.
Funding trends: assessing the efficiency of spending on the
school nutrition programme, 1995/96 to 2005/06
Figure 3 provides information on real trends in budgeted expenditure on the
school nutrition programme for the period 1995/96 to 2007/08.
Rand Values
Figure 3: Real trends in budgeted expenditure on the school nutrition
programme, 1995/96 to 2007/08 (2000 Rands)
900000
800000
700000
600000
500000
400000
300000
200000
100000
0
20
20
20
20
19
19
19
07
05
03
01
99
97
95
Financial Years
Source: Personal communication with departments of Health and Education in
2004 and 2005; Budget Review 2005
Figure 3 indicates the strong negative declines in the early days of the school
nutrition programme, and also the strong real increases that took place from
2003/04. There are varying interpretations about the status and meaning of
the expenditure patterns in Figure 3 above. Brand (2004) argued that the
declines need not be taken as a sign of lack of prioritisation of the school
nutrition programme. They rather represented a situation where targeting
had become increasingly effective and where the results of such precise
targeting led to “savings.” This is a remarkable conclusion especially in light
of the problems that government experienced with the administration of
conditional grants during this period (see public finance section). The Child
22
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Health Unit (1997) traced the origin and meaning of the reduction in
spending to the following context (page 90);
However, over the last three years, there has been a net reduction in the amount of
funds allocated to the PSNP. It was for this reason that the recommended percentage
RDA for energy in the school snack was reduced from 30 % to 20-25%.”
If we leap forward to the time when the school nutrition programme was
supposedly more effective in targeting relatively smaller learner numbers,
then one would have expected the original % RDA content to have been
restored. In addition, if these declines were in fact real savings, then the
question must be asked why national and provincial health departments did
not return to the original % RDA content, especially given the “surplus” of
funding at their disposal? The length of the real decline and the fact that
targeting problems were still identified as late as 2002 and 2003 (Hunter et
al., 2003), suggest that the real declines were more powerfully related to the
paucity of funding. Furthermore, in the context of our discussion of growth
patterns in overall health spending, it is clear that the school nutrition
programme was a victim of expansions in other areas of health spending.
The main reasons for weakened funding support to the school nutrition were
the containment of consumption expenditure, the moderation of social
spending, and real increases in the wage bill of health departments. In
contrast, the reasons why increased funding support have taken place since
2002/03 were the food crisis of 2002 (Terreblanche, 2002), the schooling
review of the Department of Education in 2003 (Naidoo, 2003), and the
nutrition programme’s insertion into education budgets and funding.
How well did provincial health departments spend the limited funds that
were available to them? Figure 4 provides information on real budgeted and
actual expenditure on the school nutrition programme for the period
1995/96 to 2004/05. We could not access provincial data for the 1995/962003/04 period because it had not been recorded in such a format by the
national Department of Health. The 2004/05 financial year was the first year
that education took responsibility for the nutrition programme.
Rand Values
Figure 4: Real budgeted and actual expenditure on the school nutrition
programme, 1995/96 to 2004/05 (Rands)
800000
700000
600000
500000
400000
300000
200000
100000
0
Actual
Source: Personal communication with the Departments of
Education in 2004 and 2005
Health and
Figure 4 does not present a linear view of actual spending, which is
premised on the idea of weak spending at the inception stage and
23
05
Budgeted
20
04
20
03
20
02
20
01
20
00
20
99
19
98
19
97
19
96
19
95
19
94
19
Financial Years
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
progressively better spending rates at each successive year. We note that the
lowest level of real actual expenditure was recorded in 1996/97. Immediately
after the low of 1996/97, the next financial year saw considerable
improvements in the spending ratio, only to be followed by larger deviations
in the following two years. Private consultants were brought in during 1995
and 1996 to help improve actual spending (Business Day, 1997), and
according to the same source, improved delivery resulted from their
interventions. The net effect of their interventions would perhaps have been
clearer if we had access to provincial actual spending because the aggregate
numbers above present a much more complicated story. The larger real
variation in actual expenditure as opposed to real budgeted allocations
happened because of variable actual spending. Significantly lower actual
spending in one year, which is followed by stronger actual spending in the
following year, may visually communicate improvements, but such "positive"
changes are a product of poor spending performance in the previous year.
From the 2002/03 financial year, budgeted and actual expenditure grew
closer together to such an extent that one could no longer discern two
different curves. It is ironic that such better spending ratios were taking
place in contexts where the number of beneficiaries had grown. This
suggests that the basic infrastructure (financial and otherwise) to implement
a school feeding programme seems to be in place. This also suggests that
larger number of poor learners may not necessarily present future underspending problems, but may in fact lead to "positive" over-spending.
Table 4 sets out the actual spending ratios in nominal terms and confirms
the trends that we observed in Figure 4.
Table 4: Actual expenditure as a percentage of budgeted expenditure on
the school nutrition programme, 1995/96 to 2004/05
Budget
allocation Actual
Expenditure
Financial Year
(Rands)
(Rands)
1995
500000
375000
1996
496000
325621
1997
496000
399376
1998
465941
351559
1999
457945
356145
2000
433401
423303
2001
458938
385659
2002
469465
449692
2003
686935
663196
2004
832200
819815
Source: Personal communication with the Departments
Education in 2004 and 2005
Actual Exp. as % of
Budget allocation
75.0
65.6
80.5
75.5
77.8
97.7
84.0
95.8
96.5
98.5
of Health and
The non-linear actual expenditure patterns that we referred to are evident in
the 1995-2001 data. Actual expenditure declined from 75 per cent in
1995/96 to approximately 66 per cent in 1996/97, only to go up to 81 per
cent the following year. These patterns are sustained right until the 2001/02
financial year. Thereafter, especially in the context of food inflation in 2002,
actual expenditure began tracking budgeted expenditure with greater
24
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
precision. The transition from the health departments to education
departments in 2004/05 did not affect overall actual spending. One of the
reasons for the successful transition was the amount of collaboration
between health and education departments in the run-up to the handover.
The handover of the school nutrition programme from health to education
means that we are now able to access provincial actual spending ratios.
Figure 5 provides information on actual expenditure on the school nutrition
programme in the first year under the control of education departments.
l
io
na
pe
Ca
Na
t
No
rth
W
es
ter
n
W
es
t
pe
Ca
er
n
No
rth
pu
m
ala
ng
a
po
M
Li
m
po
ten
g
Kw
aZ
ul
uNa
ta
l
Ga
u
te
Fr
ee
St
a
ter
n
Ca
p
e
140.0%
120.0%
100.0%
80.0%
60.0%
40.0%
20.0%
0.0%
Ea
s
Actual percentage spending
Figure 5: Actual expenditure as a percentage of budgeted expenditure
on the school nutrition programme by provincial education
department, 2004/05 (See table in the appendix)
Provincial education departments
Source: Personal communication with the national Department of
Education 2005
Figure 5 shows that nationally, provincial education departments spent on
average 99 per cent of their school nutrition allocations. There are three
exceptions to the overall pattern, namely KwaZulu Natal, Limpopo, and the
North West province. KwaZulu Natal recorded the lowest actual spending
ratio (81.8%), which is approximately 20 percentage points lower than the
national average. North West managed to spend 88.6 per cent of its
allocation, which represents a spending ratio that is approximately 12
percentage points lower than the national average. In the case of the
Limpopo province, indications are that the province spent 127 per cent of its
original allocation. While the latter figure does appear unusual, this situation
is likely to become the norm as provincial education departments revise their
targeting lists and new schools are added to the beneficiaries list. We feel
supported in this conclusion by the fact that in 2004/05, six of the nine
provincial education departments spent more than 95 per cent of their
original allocations. The increases in spending ratios are also supported by
the continuation of tender and other service delivery arrangements that were
formed under provincial health departments. While many of these contracts
were subsequently cancelled, it afforded provinces continuity in service
delivery, thus contributing to the high spending rates in 2004/05.
The 2004/05 expenditure data provide further context to two questions: the
extension of the school nutrition programme to more schools (and in
particular secondary schools) and the importance of discretionary funding.
In the context of debates about who should be included in the list of
beneficiaries, it is unlikely that the outcome of such debates would lead to
25
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
smaller number of targeted learners. In fact, with the financial expansion of
the school nutrition programme a given, beneficiary numbers will only
stabilise in the next three to five years. Four provincial education
departments have already raised concerns about the adequacy of the school
nutrition allocation. Three of these provinces would be classified as
traditionally poor, and this reinforces our suspicion about the net impact of
the Department of Education’s re-definition of poor learners on poor
provinces. The inclusion of more learners requires larger financial outlays,
which may not be available. This focuses our attention on discretionary
funding from provincial education departments' own line budgets.
Table 5 provides information for four provinces about additional funding
from their own coffers for the school nutrition programme in 2004 and 2005.
Table 5: Discretionary funding made available to the school nutrition
programme for selected provincial education departments, 2004/05
and 2005/06
As a % of
grant
allocation in
Province
2004/05
2004/05
Gauteng
R1.2 million
1.6%
KwaZulu Natal
None
N/A
Limpopo
R20 million
13.1%
Northern Cape
R3.5 million
15.6%
Source: Personal communication with
2004 and 2005
As a % of
grant
allocation in
2005/06
2005/06
R1.27 million
1.54%
R55.4 million
27.8%
Yes, but unable to verify amount N/A
11,050 million
44.9%
provincial education departments in
Five provincial education departments do not make provision for additional
funding from their own line budgets. Those that do provide additional
funding represent both poor and rich education departments. Table 5 shows
that the level of discretionary provisioning strongly differs across provinces.
In 2004/05, discretionary funding constituted approximately 2.0 per cent of
total grant funding for Gauteng, while the percentage mark in Northern Cape
was approximately 16.0 per cent. Similarly, in 2005/06, the Northern Cape's
discretionary allocation is almost half of their total grant funding, while
KwaZulu Natal invested 28.0 per cent from their own coffers. Gauteng
maintained approximately the same ratio as in the previous financial year.
The information in Table 5 suggests that while provinces are planning
extensions to the school nutrition programme, they do not have the funding
bases to support expansion. Comments about the rapid extension of the
school nutrition programme should therefore be regarded with great caution.
Based on the funding information in Table 5, if provinces are committed to
prudent spending on their conditional grant allocations, then there will not
be a rush to add substantial numbers of new beneficiaries. However, the
enduring popularity of the programme, coupled with political pressure, may
force departments to expand participation in the school feeding scheme. This
could lead to the same over-spending scenario that Limpopo experienced,
and which was primarily caused by the rapid addition of new beneficiaries.
Given the pressures to deliver school nutrition to the broadest majority of
poor learners, are there significant differences in the per learner costs of
school feeding amongst provincial education departments?
26
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Figure 6 provides information about the estimated per learner costs of the
school feeding programme in 2004/05 and 2005/06.
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
Cost per learner 2004/05
ten
Kw
g
aZ
ul
u
N
at
al
Li
m
po
po
M
pu
m
ala
ng
N
a
or
th
er
n
Ca
pe
N
or
th
W
es
t
W
es
ter
n
Ca
pe
Ga
u
Ca
pe
Fr
ee
St
at
e
Cost per learner 2005/06
rn
ste
Ea
Estimated cost per learner per day
(Rands)
Figure 6: Estimated cost per learner in 2004/05 and 2005/06 by
provincial education department (See table in appendix)
Source: Personal communication with provincial education departments in
2004 and 2005
Figure 6 shows that provincial education departments maintain the same
cost structure in the two years represented above. The highest per learner
expenditure is recorded for the Free State and North West provinces (R1.50
per learner per day), while the lowest per learner average is recorded for
Mpumalanga (R0.83 in 2004/05) and KwaZulu Natal (R0.90 in 2005/06).
These are estimated actual costs and do not reflect optimal levels of
delivering school nutrition to poor learners. Given the lack of guidance in
this regard, it is difficult to judge these numbers and the distances among
the provincial education departments. Much more research is required into
this area before meaningful judgments can me made about cost-efficiency in
the delivery of provincial school feeding programmes. Figure 6 does not show
the differences in per learner costs between urban and rural areas, but this
is a fact for many provincial education departments. As per provincial
differences, we have no meaningful scale to judge an urban per learner cost
of R1.50 against R1.30 for rural learners in the North West province.
The next section continues the discussion of service delivery issues, but now
focuses more narrowly on the overall effectiveness of spending on the school
nutrition programme.
Output trends: assessing the effectiveness of spending on the
school nutrition programme, 1995/96 to 2005/06
Van der Walle (1998) notes that society stands to benefit greatly from
directing resources to people who need it most, because if we succeed, both
the individual's well-being and the collective well-being of a country are
enhanced. The benefits of targeting are therefore aimed at enhancing
government's anti-poverty campaigns and decreasing the vulnerability of the
poor. With these introductory remarks, we now proceed to an examination of
old and new targeting practices.
27
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Table 6 provides information about the targeting strategies that were
employed by provincial health departments in identifying individual and
school-level beneficiaries.
Eastern Cape
X
X
Free State
Gauteng
X
X
KwaZulu Natal
X
X
Limpopo
X
X
Mpumalanga
X
X
Northern Cape
X
North West
X
Western Cape
X
X
X
X
Criteria
Individual
Grade/Age
School
Province
Geographical
Table 6: Provincial health targeting strategies in 2001/02
Grade R to Grade 4. All farm schools.
Walking distances to school more than 3 KM
Caregivers that cannot provide for learners.
Social problems
Availability of running water at school
Availability of sanitation at school
Availability of electricity at schools
X
Rural schools
Disadvantaged schools in townships
Farm schools
X
Rural, semi-urban, farm or other informal settlements
Poverty gap
Per capita income of communities
X
All rural and farm schools
Rural schools-100%learners in schools
Urban-10% of learners in schools
Non-urban-80% of learners in schools
X
Grade R to Grade 3
Source: Louw et al. (2001: 49 and 50)
The provincial health targeting strategies depicted above were not the only
methods that identified poor primary school learners. The identity of poor
public primary schools was determined through the school funding norms
(education) and the diverse provincial health targeting strategies (school
nutrition) displayed in Table 6 above. Even informed sources that discussed
the targeting process for the school nutrition programme (Louw et al., 2001)
did not explicitly refer to the then emerging school funding norms. This
creates the perception that there was no recognisable poverty targeting
strategy in provincial education, yet in 2001, the school funding norms
entered its second year of implementation. This situation reflects poor interdepartmental communication in the implementation of a policy that cuts
across the education and health sectors. In fairness to provincial health
departments, it must be said that there was no uniform targeting strategy
across provincial education departments and therefore little confidence that
the provincial education targeting process would accurately identify the
poorest of the poor.
The common denominator that runs through the provincial targeting
strategies in table 6 is the (official policy) recognition that schools in rural
and semi-urban areas, farm schools, and schools in informal settlements
should enjoy the highest priority.
Besides this minimum consensus,
provincial health departments developed their own indicators to identify
participating and needy schools and learners. Table 6 makes the point
28
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
clearly that the school nutrition programme at the time was relatively
undeveloped, because some provinces such as the Eastern and Western
Cape only targeted the foundation phase. While such targeting would no
longer be acceptable today, in the context of what the Child Health Unit
(1997) argued for, such targeting would have made perfect sense then. Six
provinces also indicated that they used some form of individual targeting,
i.e. providing only to particular children in schools, rather than to all
children in targeted schools. Mtyala (2003) wrote about the Western Cape
context and pointed to the socially undesirable impact of individual
targeting, especially the stigmatisation of children who receive these meals.
The dual system of identifying poor public primary schools through
provincial health and provincial education targeting strategies resulted in
targeting systems that did not necessarily identify the same schools as
“disadvantaged.” Given the government’s rallying call of “integrated service
delivery”, this non-coincidence reduced the overall impact of redress
spending for many schools. This was the case because the ideal must have
been that those schools or learners who were participating in the school
feeding scheme should also have received affirmation through the school
funding norms. We are unable to provide any information about the number
of such mis-classified schools, but the fact that provincial education
departments are presently revising their databases suggests that such
mistakes were common.
Table 7 provides information about provincial education targeting practices
in 2004/05.
29
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Eastern Cape
Grade R to Grade 7
Quintiles 1 and 2
Free State
Grade R to Grade 7
Quintiles 1 and 2
Gauteng
Grade R to Grade 7
Quintiles 1 and
Quintiles 3 and 4
KwaZulu Natal
Limpopo
Grade R to Grade 7
Grade R to Grade 7
Mpumalanga
Grade R to Grade 7
Northern Cape
Grade R to Grade 7
Grade R-Grade 4 or
Grade 1- Grade 5
North West
Western Cape
Grade R to Grade 7
Additional
features
Priority
Quintiles
Grades
Province
Table 7: Summary features of targeting practices in 2004/05 by
provincial education departments
For non-farm schools, Grade R-4.
For farm schools, Grades R to 7
are targeted
School funding norms were used
to update inherited provincial
health databases
2;
Quintiles 1 and 2,
2038 schools
Quintiles 1 to 3
Quintile ranking was
problematic. Physical
condition of buildings
was considered instead
of poverty index.
100% coverage for quintiles 1 and
2, while 75% coverage for rest
1420 schools located in quintiles
3-5
inherited
from
health
department
All farm schools are targeted.
All farm schools are targeted.
Quintiles 1 to 5
Quintile 1 and 2
Schools within rural and urban
presidential nodes are top priority.
We feed the first five grades
Poverty Index targeting
the
four
poorest
segments
Poorest
(0.9=90%),
second
poorest (0.8=40%) and third and
fourth [combined] poorest (0.7 +
0.6 =25%)
Source: Personal communication with provincial education departments in
2004 and 2005
In interpreting targeting trends in table 7, we have to remember that the
Department of Education's proposed re-definition of poor learners does not
apply. Latest indications are that provinces are still in discussion with the
Department of Education and that the process will only be finalised towards
the beginning of 2006.
In contrast to the information in table 6, table 7 displays far greater
uniformity in the targeting strategies of provincial education departments.
Across provinces, there was near uniformity in the selection of Grade R to
Grade 7 as the main beneficiaries. Early reviews of the school nutrition
programme had consistently emphasised the policy and nutritional
importance of focusing on the very young. Researchers who had been calling
for such changes would take great heart from the initiative to introduce
Grade R as part of the participating group of learners benefiting from school
feeding. However, our optimism of this strategy is tempered by the fact that
government-supported Grade R facilities are a drop in the ocean compared
to the numerous ECD facilities that do not receive funding support from the
State. Bolowana (2004) noted that in KwaZulu Natal, many parents rushed
their children out of un-subsidised ECD programmes into primary schools to
access policies of fee exemptions and of course the school nutrition
programme. Bolowana (2004) further noted that in rural and poverty nodal
areas, very few pre-schools were either registered or subsidised and are
consequently dependent on fees by parents.
30
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
There are two exceptions to the strategy of targeting Grade R to Grade 7,
namely the Eastern Cape and North West. Both offer a truncated version
and only target learners from Grade 1 to Grade 4 in the Eastern Cape, and
Grade R to Grade 4 in the case of North West. In the case of the Eastern
Cape, budget constraints have forced a strategy where only Grades 1 to 4 in
poor non-farm schools are targeted, while Grade 1 to Grade 7 are targeted in
farm schools.
Apart from the opportunity costs of not feeding learners in
Grade R, the Eastern Cape’s non-farm schools are poorly covered. In
addition, the number of feeding days varies between three and five days, but
recently, the Eastern Cape Department made it clear that it can only afford
to feed children three days per week (Cape Times, 2005). In North West,
similar budgetary reasons are provided for the poor coverage of the primary
school grades. In 2004/05, only poor learners in Grades R to 4 were fed.
Barring these exceptions in targeting, targeting in 2004/05 was boosted by
the selective use of poverty quintiles. While the new nationally defined
poverty quintile system has not been implemented, some departments
continued with inherited schools and added new schools based on their
quintile location. This was one of the reasons for the sharp increase in
beneficiary numbers in 2004/05. Provinces find themselves in a difficult
position, because the application of the school funding norms must mean
that many of these schools should be removed from the list of participating
schools. In fact, we argue that such arrangements make it hard to extend
the benefits of the school nutrition programme to more poor primary and
secondary learners. This situation provides strong evidence of targeting
inefficiencies and confirms the non-coincidence of old and new targeting
strategies in identifying poor primary schools. KwaZulu Natal offers a strong
example of this dilemma because it has a sizeable number of beneficiaries in
the more “affluent” poverty quintiles. From a preliminary analysis of the old
and new targeting practices, it would appear that the education funding
norms (imperfect as they are) provide more precise targeting of poor schools.
In the traditionally rich province of Gauteng, this department has chosen to
affirm learners in poverty quintiles 1 to 4, even though the degree of
coverage differs slightly across the quintiles. Although the distribution of the
Western Cape cannot be broken down into quintiles, it is almost certain that
should a quintile concept be imposed, we would find that learners in
“quintiles” 1 to 4 are affirmed. However, table 7 makes it clear that the
difference in coverage rates for poor learners in different poverty segments is
substantial.
Targeting strategies in 2004/05 reveal an uneasy alliance of grade targeting
and the selective and non-uniform use of poverty quintiles. Should
developments concerning the proposed provincial poverty quintiles be
finalised, earlier comments about the differential impact of such schemes on
rich and poor provinces still apply. What were the fruits of targeting and did
targeting succeed in reaching a larger number of learners and schools?
Figure 7 provides information on the number of schools that were reached
as a proportion of targets set by provincial health departments.
31
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
25000
1.2
20000
1.0
0.8
15000
0.6
10000
0.4
5000
0.2
0
Proportion of target
reached
Total number of schools
targeted
Figure 7: Total number of school targeted and number of schools
reached as a proportion of targets, 1994/95 to 2003/04 (See table in
the appendix)
Targeted
Reached
0.0
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
Financial Years
Source: Personal communication with Departments of Health and Education in
2004 and 2005
There are three main trends in the original targeting data: an initial surge at
the start of the programme; followed by strong declines over the 1997/98 to
1999/00 period; and finally, a relatively slow expansion of total number of
participating schools between 2000 and 2003. Between 1995/96 and
1996/97, there was a decline of 3085 schools participating in the nutrition
programme, which represents a 15.3 per cent decrease. Thereafter, for three
successive years (1997/98-1999/00), the number of schools declined by
1858 or 10.3 per cent. Since 2000/01, there has been a steady, but slow
increase in the number of new schools that were added to the feeding
scheme. In the four-year period (2000/01-2003/04), the number of
participating schools increased from 16200 to 17000. This represents a 5.0
per cent increase, which in comparison with the multiple declines in earlier
years, is relatively small. The net effect of these trends is that when we
compare the total number of schools that participated in 1995/96 (20110) to
that in 2003/04 (17000), 3110 fewer schools (or 15.4% less) participated in
2003/04 compared to 1995/96.
We would expect the number of participating schools to increase, especially
in view of the directive from the Department of Education that schools
inherited from provincial health departments must be retained. However, the
rate of increase would differ according to the fiscal capacity of provinces to
absorb the added costs of greater coverage. We do not support the view that
the decline in number of schools is related to better and more effective
strategies. In the overall context of declining and weak financial support for
the school nutrition programme especially during the nutrition programme’s
supervision under health, the decline is directly related to availability of
funding. At the inception of the programme, there was clearly an attempt to
cover as many primary schools as possible.
At the start of the school nutrition programme, targeting achieved variable
success rates. During this same period (1994/95-1999/00), there was no
linear relationship between the total number of participating schools and
target success rates. We see for example that there was a net increase in
participating schools in 1997/98, yet the targeting success rate increased by
32
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
approximately 4.0 per cent. It is only after 2000, when the number of
participating schools was allowed to increase slowly, that some stability had
been reached concerning targeting success rates. In addition, the slow
increase in the number of participating schools did not lead to a dramatic
decline in the targeting success rates. This provides further evidence that
smaller number of participating schools did not necessarily mean higher
targeting success rates. In fact, the data show the opposite.
The number of schools that are successfully reached should not be
exclusively construed as evidence of policy success. The fact that school
feeding did reach schools does not reflect on the type of services that was
delivered, the quality of the food, or the efficiency with which these services
were delivered. Strong parallels can be drawn with the ECD sector, where
policy success is measured as the total number of new governmentsponsored Grade R sites. It is relatively easy to establish a Grade R site, but
it is much more difficult to ensure that quality and adequate education is
delivered at the site. Furthermore, in the context of pre-2004 targeting
strategies, reaching a school was synonymous with working with only a few
grades so that large numbers of learners would not have been serviced. The
opposite also happened where in spite of limited targeting mandates in
schools, all learners were fed, which resulted in the dilution of portions
(Louw et al., 2001).
Figure 8 provides information about the total number of learners that were
targeted and the success rates in reaching targeted learners for the period
1994/95 to 2004/05.
8000000
1.2
7000000
1.0
6000000
0.8
5000000
4000000
0.6
3000000
0.4
2000000
0.2
1000000
0
Proportion of target
reached
Total number of learners
targeted
Figure 8: Total number of learners targeted and number of learners
reached as a proportion of targets, 1994/95 to 2004/05 (See table in
the appendix)
Targeted
Reached
0.0
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Financial Years
Source: Personal communication with Departments of Health and Education in
2004 and 2005
Unlike the multiple trends that we observed in the total number of schools
participating in school feeding, learner numbers appear one-dimensional.
Between 1995/96 and 2003/04, learner numbers dropped from a “high” of
approximately 6.8 million in 1995/96 to 4.8 million in 2003/04. This
represents a percentage drop of 29 per cent. If we couple this to less than
perfect targeting success, especially for the period prior to 2000, then the
true impact of the school nutrition programme appears very limited. It is
only after 2000 that the number of those reached as a proportion of original
33
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
targets, consistently registered tallies above 90 per cent. This success takes
place against a background of reductions in the total number of learners
that were targeted. This trend appeared to have developed since the 2002/02
financial year. The extent of the backlogs that will have to be faced now is
demonstrated with the learner data for 2004/05. More learners have
participated in the school feeding scheme than was planned (107%), and we
expect much the same to happen in the next few years.
Table 8 further contextualises the learner numbers by comparing
participating learner numbers in the school nutrition programme to the total
primary school learner population. We are now able to do this for the first
time for each provincial education department.
Table 8: Primary school learners reached as a proportion of total
primary school learners by provincial education departments in
2004/05 and 2005/06
Province
Eastern Cape
Free State
Gauteng
KwaZulu Natal
Limpopo
Mpumalanga
Northern Cape
North West
2004 Absolute
Number
948 574
149 710
325 036
1 338 337
1 157 193
491 362
121 000
491 068
2004 as a % of
primary
learners
56%
39%
33%
80%
92%
89,2%
88 %
37%
2005 Absolute
Number
983 412
234 534
400 866
1 371 777
1 140 521
492 687
124 446
380 385
2005 as a % of
primary
learners
58%
62.5%
40%
84.1%
99%
90,3%
92 %
62%
Western Cape
149 304
25.6%
156 554
27.4%
Source: Personal communication with provincial education departments 2004
and 2005
Table 8 shows two interesting features, namely a progressive increase in the
coverage rate of primary school learners, and large cross-provincial
differences in coverage rates of primary school learners. All the provincial
education departments exhibit progressive increases in coverage of primary
school learners from 2004/05 to 2005/06. Free State increased its coverage
rates from 39 per cent in 2004/05 to 62.5 per cent in 2005/06, while
Gauteng moved from 33.0 per cent to 40.0 per cent over the same period.
Limpopo appears to have moved to universal coverage in 2005/06 (99%),
while Mpumalanga and the Northern Cape achieved coverage rates above
90.0 per cent in the same year. The range of coverage of primary school
learners is large and varies from 27.4 per cent in the Western Cape in
2005/06 to 99.0 per cent in Limpopo. Historical targeting practices are one
of the reasons for this divergence, because in a province such as the Western
Cape, restricted grade targeting (Grade R to grade 3) was practised.
Table 8 further contextualises the extension of the programme into poor
secondary schools. In Limpopo, available resources have been exceeded in
order to achieve near universal coverage of primary school learners. The
fiscal space for expansion to poor secondary schools does not exist and the
servicing of a new group of schools would require additional resources.
However, in provinces such as the Eastern Cape and North West, expansion
34
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
of the number of participating poor primary schools is likely to continue, and
therefore any expansion into secondary schools appears remote.
Our final service delivery information relates to the frequency of feeding.
Frequency of feeding had been identified by all the major reviews as one of
the biggest challenges. Table 9 provides information about the frequency of
feeding per day, per week, and over the entire school calendar year for each
provincial education department.
Table 9: Total number of feeding days in 2004/05 and 2005/06 by
provincial education department
Total number
Total number of
planned
No. of times of Frequency
per of feeding days feeding days in
Province
feeding per day
week
in 2004
2005
Eastern Cape
Once
3-5 days
156
156
Free State
Once
5 days
186
184
Gauteng
Once
5 days
142
192
KwaZulu Natal
Once
5 days
156
173
Limpopo
Once
3-5days
156
156
Mpumalanga
Once
5 days
156
156
Northern Cape
Once
5 days
156
156
Once
4 days
156
156
North West
Western Cape
Once
5 days
170
170
Source: Personal communication with provincial education departments in
2004 and 2005
The results in table 9 follow recommendations about the standardisation of
the minimum number of feeding days. Most provinces feed learners five
times per week and conform to the minimum number of 156 days per
academic calendar year. Eastern Cape, Limpopo and North West deviate
from the strict 5 days per week feeding schedule and offer variable feeding
days. In these provinces, deviation from the 5 times per week feeding
schedule is caused by budgetary constraints. In the Eastern Cape and
Limpopo, it would appear that the 5 days per week feeding schedules are
normally available at the beginning of each financial year. However, such
schedules are modified in later parts of the year, although communities are
notified right from the start about future changes to the feeding schedule.
In 2004/05, two provinces, namely Free State (186 days) and the Western
Cape (170 days) provided more feeding days than the prescribed minimum of
156 days. Gauteng was the only province that fed learners over a lower
number of days (142 days). In 2005/06, projections of the total number of
feeding days suggest that three provinces (Free State, Gauteng, and Western
Cape) have planned to feed learners over a period that exceeds the minimum
required number of feeding days. The fact that most provinces stick to the
minimum number of feeding days is an indication of the tightness of funds
in a context where increasing demands will be made upon their school
nutrition budgets.
35
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Summary remarks about the efficiency and effectiveness of the
school nutrition programme, 1995/96 to 2005/06
We identify six main points from the review of the efficiency and effectiveness
of spending on the school nutrition programme between the period 1995/96
to 2004/05:






Under the supervision of the Department of Health, the school
nutrition programme was subjected to real declines since its
inception. Discounting under-expenditure, spending on the school
nutrition programme recovered only after 2001. The declines are not
due to better targeting or efficiency gains, but reflect poor funding
prioritisation of a very important public policy.
Highly variable and inefficient actual spending patterns characterised
the programme between its inception and 2001/02. This directly
reduced the overall effectiveness of the programme in serving poor
learner communities.
The absolute number of participating schools decreased by
approximately 3100 between 1995/96 and 2003/04, which
represents a 15.4 per cent decline. When the number of targeted
schools did increase, the margin was small and indicated an increase
from 16 200 schools in 2000/01 to 17 000 in 2003/04.
Learner participation dropped from 6.8 million learners in 1995/96 to
approximately 4.8 million learners in 2003, which represents a
reduction of 29 per cent.
High expenditure ratios in 2004/05 appear to reflect growth in
beneficiary numbers, which is partly the result of inherited targeting
inefficiencies and partly due to revisions to inherited databases by
provincial education departments.
Differences in targeting practices between education and health prior
to 2004/05 caused the same primary school to be treated differently
by health and education poverty targeting strategies. This weakened
the overall impact of redress spending and reflected poor interdepartmental communication and co-operation.
SECTION 4: CONCLUDING REMARKS
The school nutrition programme can rightly be regarded as a “survivor.” Its
roots are traced back to the inception of the new democratic government in
1994 and it is now guaranteed a public financial life beyond 2008/09. This
remarkable longevity hides the important story of a bruised public
programme, which for the moment, appears to undergo rehabilitation under
the supervision of the Department of Education.
The vital question about the conceptual and classificatory status of the
school nutrition programme appears to have been settled. This had not
always been the case because at its inception, there were strong reservations
about a school feeding programme that carried the name "school nutrition
programme." Researchers felt that the primary school nutrition programme
was using scarce State resources and its stand-alone status as a feeding
programme could not be justified. Crucial interventions such as mass deworming medication, the delivery of micronutrients, control of parasitic
worm infections, family planning, life skills, and education aimed at
36
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
reducing tobacco and alcohol use were squeezed out by the expensive and
logistically intractable school feeding programme. Researchers continued to
resist calling the school nutrition programme a “comprehensive nutrition”
programme and its designation as an exclusively school feeding programme
was conceptually complete in the post-2000 period. The food inflation crisis
of 2002 and Cabinet’s strong focus on anti-poverty programmes brought the
school nutrition programme under the ambit of anti-poverty programmes. In
a rigorous way, none of the costs and policy objections that were raised
against the school nutrition programme had become invalid. However, the
definition of the school nutrition programme as an anti-poverty programme
removes some of the reasons that would have restricted school feeding and
made narrower targeting the modus operandi. Broad based affirmation of
poor learners has now become a necessity, in spite of the limitations of the
State’s overall anti-poverty programmes. The relatively large real increases in
spending on the school nutrition programme over the present MTEF confirm
the status of the programme as an anti-poverty measure.
So what were the public finance contexts that the school nutrition
programme had to navigate and how was it affected by each of these
contexts? We reviewed three public finance contexts, namely the period
immediately after the 1994 elections, the period between 1996 and 2000,
and the present MTEF (2004-2007). The immediate challenges following the
national elections in 1994 were two-fold: the State had to provide greater
opportunities to the historically marginalised, while simultaneously dealing
with the public finance excesses of the previous regime. Key to the
successful transition was the re-prioritisation of public expenditure away
from consumption expenditure to development and capital expenditure. This
requirement made it difficult, if not impossible, to start new projects by
increasing existing staff levels. These developments were the roots of the
present conditional grant framework, which does not conceptualise
additional staff as a necessary condition for the successful implementation
of grant funding projects. Scarce human talent was therefore over-committed
and locked into programmes whose challenges were not clearly understood
at the outset. Added to this staff burden was the extension of the principle of
“no real growth” in funding for earmarked projects. For the school nutrition
programme, the net effect of these actions was a significant drop in the total
number of participating schools and learners. It was only poor actual
expenditure that prevented a full-scale exposure of the inadequacies of this
programme.
The school nutrition programme’s next destination was its integration into
the Department of Health’s integrated nutrition strategy. This period lasted
from 1998/99 to 2003/04 and included two distinct phases in the evolution
of public finances in South Africa. The phase that ran from 1996 to 2000
was characterised by government’s macro-economic strategy (Gear)
informing fiscal policy and budget policy. Gear’s main influence was on
moderating social services’ spending on the national purse. The post-2000
phase saw a mild expansionary fiscal stance adopted by the South African
government. While real reductions to the school nutrition allocations for the
period 1996-2000 are unsurprising, we were surprised by the results of the
real growth after 2000. Consistent with social service programmes’
reductions in the 1996-2000 period, the school nutrition programme was
reduced in real average annual terms by 9.7 per cent, while actual
37
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
expenditure decreased by 3.9 per cent over the same period. The reason for
the smaller real loss in actual expenditure is because of variable spending
patterns during this period. Although the real average annual growth rate
between 2000 and 2003 was positive (10.3%), this was mainly due to the
large increase in the allocation in 2003 (38%). For all the other years in this
period, budgeted school nutrition programme funding continued to shed real
resources. This was in sharp contrast to consolidated provincial health
spending, which appeared to have recovered well after 2000. Higher wage
costs in provincial health are given as one of the main reasons why the
school nutrition allocation and the integrated nutrition allocation as a whole
did not benefit.
The third destination of the school nutrition programme was its location as a
national conditional grant under the supervision of the Department of
Education. The education sector took over funding and supervisory
responsibility in 2004/05, and its inheritance was marked by two features:
school nutrition allocations that were declining in real value since its
inception, and a conditional grant framework that was much more enabling
in promoting effective and efficient implementation. The school nutrition
programme will change in two important ways. The first relates to its
insertion in education budgets that have grown slowly, but consistently, in
the post-2000 period. We have argued that this consistent growth was
targeted at meeting a policy framework that allowed activities and
expenditure categories of relatively small magnitude to grow in real terms.
The school nutrition programme fits the bill and is projected to grow by a
real average annual rate of 6.4 per cent over the 2004 to 2007 period. It
would appear that the funding neglect that this programme suffered under
the supervision of the Department of Health is set to come to an end.
The second change relates to the total number of schools and learners that
are earmarked to benefit from this programme. We showed how faced with
real reductions in resources, the school nutrition programme in its earlier
guises suffered blows in terms of the reduction of participating schools and
learners. Now, with real average annual growth rates guaranteed, and
targeting processes re-visited, the total number of schools and learners
across provinces is projected to increase. However, the re-definition of poor
learners according to national poverty quintiles has different implications for
rich and poor provinces. In the case of the latter, larger number of poor
learners would define the poorest of the poor and this would increase the
funding costs for poor provinces. Rich provinces would have fewer learners
in the poorest brackets and would have more space to target a diversity of
poor learners. If we add the directive from the national Department of
Education that no participating schools inherited from the Department of
Health lists should be removed, then it is easy to understand why Limpopo
spent 127 per cent of its original allocation in 20004/05. We predicted that
the next few years will be used to finalise beneficiary numbers and this
raises important questions about the adequacy of present allocations and
the need for additional funding from provincial education departments.
In assessing the overall efficiency of spending, we found that the school
nutrition programme exhibited highly variable spending levels from its
inception until the end of 2001. In addition, real levels of spending on this
programme showed a consistent negative decline since its inception. It was
38
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
only at the start of the 2003/04 financial year that spending recovered and
grew progressively since then. The combination of variable actual spending
and low funding priority had a negative impact on overall service delivery
and quality. It was not the case that the "little" had been spent well, but it
was rather that the barest little could not be spent with all the other
accompanying service delivery problems. We linked poor financial servicing
of the school nutrition programme to declining number of participating
schools and learners. Similar to the patterns on the financial allocations, the
total number of participating schools and learners only started growing
again in the post-2000 period. In addition, this growth was relatively small
compared to the reductions in the number of schools and learners in earlier
periods. We rejected interpretations that suggest that decreases in the size
of the school nutrition budget and total number of schools and learners
reflect increasingly effective targeting strategies. In fact, we critically
analysed the targeting data and found many inefficiencies.
Arguably, the most telling story from the time when the Department of
Health had supervisory responsibility for the school nutrition programme
concerns the complete negation of the school funding norms. While the
school funding norms were only introduced in 2000, it nonetheless started a
process (flawed as it was), of identifying the most needy schools. However,
provincial health targeting proceeded without knowledge of the school
funding norms and the same can be said about provincial education
planners' ignorance of health targeting strategies. This led to a situation
where a dual targeting strategy did not necessarily identify the same schools
and learners as "needy." Ideally speaking, a school where learners are fed
should also have been part of the poorest poverty quintiles. These two types
of spending are complementary and the fact that the two targeting systems
did not always identify the same schools as poor reduced the effectiveness of
spending on anti-poverty programmes. Targeting was further compromised
because of the diversity of targeting strategies and the fact that some of
these strategies produced socially undesirable consequences.
The inefficiency of the old targeting system is only now being addressed
because provincial education departments found many schools that were
incorrectly classified or databases that were outdated. This revision
produces spending consequences that are not always within the reach of
provincial education departments. In the first year of the school nutrition
programme's operation under the command of the Department of Education,
provincial education departments managed to spend 99.0 per cent of their
original allocations. This is an extremely high expenditure spending ratio
and in the case of Limpopo, this ratio stood at 127 per cent. This was
indicative of a process of re-visiting databases and increasing the number of
schools and learners. In some cases, the programme would simply be
extended to more learners in the same schools, but the net effect is an
enlargement of the funding commitments of provinces. Our examination of
provincial education funding discretionary bases showed that most
education departments do not have the wherewithal to finance a rapid
expansion of learner numbers. This suggests that the overall envelope of the
school nutrition programme needs to be increased or provinces have to brace
themselves for continued over-spending of conditional grant funding.
39
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
So how does the story of the school nutrition programme continue? The
immediate future of the school nutrition programme is strongly related to
the re-designing of the school funding norms. Delays in the legislation
process concerning the school funding norms presently prevent the fullblown re-definition of the school nutrition programme from a grade-level
focus towards a broad anti-poverty focus. However, even if new school
funding norms legislation can be enacted, concerns remain about the
possible fragmentation of implementation across provinces. This suggests a
serious re-think about how changes in school funding legislation deepen
existing implementation fault lines.
___________________________________________________________________________
NOTES
We would like to thank Ms Debbie Budlender, a senior researcher with the Community
Agency for Social Enquiry (CASE), for her thorough review of the first draft version of this
paper. We would also like to thank the nine provincial co-ordinators for the valuable
information they provided. We relied a lot on their assistance in completing information
on key tables and graphs.
i
Russell Wildeman is an Education Specialist with the Budget Information Service
(Idasa) and Nobuntu Mbebetho is a Research Assistant with the same organisation.
ii
In the Cabinet briefing that signalled the handover of school nutrition from the
health to the education sector, the Cabinet statement read: "Cabinet was briefed on
a thorough review of the Primary School Nutrition Programme and welcomed
progress that has been made in ensuring food security among primary school
children in poor communities. In order to further improve efficiency in the
Programme, Cabinet decided as follows:
iii




the primary target of school feeding should also include Grade R learners in
targeted primary schools;
funding for the Programme should be increased annually in a phased
manner to cover all areas in need, in accordance with standardised menu
options;
the programme should be transferred from the Department of Health to the
Department of Education by April 2004 given the Programme's educational
outcomes and the fact that the Department is functionally responsible for
schools; and
work should continue to improve the efficiency of the Programme, including
standardised menu options and monitoring system as well as increased
involvement of School Governing Bodies and women's groups" (Cabinet
Statement, 18 September 2002, our own emphasis).
Furthermore, in response to the Congress of South African Trade Unions' section 77
notices regarding job losses, poverty and restructuring of State assets, Cabinet
responded: "The consolidation and extension of the School Nutritional Programme is
itself a concrete demonstration of the Government's preoccupation with the issues
affecting the standard of living of our citizens" (Cabinet Statement, 20 September 2002).
40
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Arguably, the strongest review of the primary school nutrition programme had been
carried out by the Child Health Unit at the University of Cape Town. The strength of their
report concerns the willingness to identify both policy benefits and policy costs and then
arguing the case for the retention or rejection of the programme. The unit was extremely
critical of the school nutrition programme as an imposed feeding programme and for its
inability to find resources for what these researchers considered "vital" activities. The
counter to a vertically imposed feeding programme is: "As part of the Integrated Nutrition
Programme, the PSNP should become a spring-board for the development of
community-based nutrition programmes. Given the number and distribution of schools
(many more than health facilities), and given the promotion of parental involvement in
schools (as part of the South African Schools Act), a transformed PSNP could become a
foundation and channel for the development of community-based nutrition activities"
(Child Health Unit, 1997: 89, our own emphasis).
iv
Our approach to the research of key areas in education funding is to keep our
arguments as simple as possible. This leads us to minimise complicated linkages and
networks of the education sector with other social sectors. We regard the task of
understanding education-specific funding issues as demanding and through our
research, we are hoping to develop a better handle on the main issues. It is only then
that we would be able to maximally extend our analyses to other areas and integrate
education concerns more successfully with other social service and service delivery
concerns. This leaves us open to a charge of "minimalism", but we defend this position
by regarding this paper as a first instalment/engagement with the issues that defined
the school nutrition programme.
v
The evidence cited in the Louw et al. (2001) study is exclusively oriented towards
supporting the usefulness and efficacy of school feeding schemes. No evidence is
reported concerning studies that cast doubt about the effectiveness and efficiency of
school feeding schemes. This is unlike the Child Health Unit (1997) study that
considered both sides of the debate.
vi
Consistent pressure from civil society and the African National Congress (ANC) led to a
series of review documents concerning the funding of public schools. At the heart of
these debates was a concern that poor learners were defined differently across provinces
in spite of similar placement in poverty quintiles. The proposed policy that attempts to
remedy these defects does not enact funding prescriptions from the Department of
Education but merely proposes various "adequacy benchmarks." These benchmarks are
figures provided by the Department of Education but have as yet not been properly
unpacked by government or civil society. These changes are proposed in Government
Notice No 26911 of 2004: Call for Comment on the Education Laws Amendment Bill, 2004.
See also, "Funding of schools: Policy issues and proposed amendments to the Norms and
Standards", a presentation to the portfolio committee on education, National Assembly,
October 2004.
vii
In the present MTEF (2004-2007), the school nutrition allocation is projected to grow
by 6.4 per cent in real average annual terms. The Division of Revenue Bill 2004 reflects
on the significance of the school nutrition in the following manner: "The Primary School
Nutrition Programme is a government programme for poverty alleviation specifically
initiated to uphold the rights of children to basic food. For this reason, there is a
national mandate to fund, spend and account transparently before government and the
public" (our emphasis). And on the life span of the grant "It is envisaged that, given the
economic climate in the country and the impact of HIV and AIDS, the need for such a
grant will persist for another 8 to 10 years at least."
viii
Section 2 of the Reconstruction and Development Programme Fund Act of 1994
establishes the various sources of the RDP Programme Fund: " There is hereby
established a fund to be known as the Reconstruction and Development Programme
Fund, which shall be credited withix
41
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
(a) money appropriated by Parliament for the fund;
(b) domestic and foreign grants;
(c) interest derived from the investment of money standing to the credit of the fund;
(d) the proceeds derived from the sale of state assets for the reconstruction and
development projects and programmes referred to in section 3; and
(e) money accruing to the fund from any other source."
From the evidence that we collected from earlier research projects, it became clear that
the availability of staff to implement conditional grant funding at the sub-national level
appears to be one of the main factors explaining actual expenditure ratios. We are only
referring to the implementation of this grant and have not even looked at the cost of
monitoring and evaluation. The latter has become crucial in the policy discourse after the
1999 national elections.
x
Over the 1997/98 to 2001/02 period, consolidated provincial education budgets
decreased at a real average annual rate of approximately 2.0 per cent, while health
budgets actually increased by 0.2 per cent over the same period. What makes the decline
in education spending so pronounced was that total provincial spending declined by 0.7
per cent on average compared to an average decline of 2.0 per cent in consolidated
education. Sources consulted were National Treasury's Intergovernmental Fiscal Review
2001 and 2003.
xi
MINMECs are institutional forums in the budget process, which consist of the national
minister in a portfolio and his/her nine MECs (member of the Executive Councils in
provincial government). Policy and policy co-ordination issues are discussed and the
overall purpose of this forum is to drive political consensus on the most important
funding and implementation challenges facing the sector.
xii
The Division of Revenue Bill 2001 (DOR Bill) institutionalised most of the discussions,
debates and improvements that were driven in the Budget Council and other budget
institutions. In the DOR Bill 2001, an entire section was dedicated to explaining a
framework for all grants and the specific criteria that were to define grant funding in
future. Unlike the Division of Revenue Bill 1998, the 2001 version specifically requested
that the following criteria be included in the consideration of grant funding: The purpose
and conditions of such allocations; the measurable outputs; reasons why the purpose of
the grant cannot be achieved through the equitable grant funding system; the projected
life span of the grant; the criteria that would guide the division of such allocations at the
sub-national level; and the recommended payment schedule.
xiii
42
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
In section 3 (1) of the South African Schools Act, No 84 of 1996, the definition of
compulsory attendance is defined as: "Subject to this Act and any applicable provincial
law, every parent must cause very learner for whom he or she is responsible to attend a
school from the first school day of the year in which such learner reaches the age of
seven years until the last school day of the year in which such learner reaches the age of
fifteen years or the ninth grade, whichever occurs first." Although the National Norms
and Standards for School Funding re-affirm these distinctions in section 19, it appears to
contradict these in section 44, which states "An important assumption underlying these
national norms is that the national and provincial levels of government will honour the
State's duty, in terms of the Constitution and the SASA, to progressively provide
resources to safeguard the right to education of all South Africans. However, educational
needs are always greater than the budgetary provision for education. To effect redress
and improve equity, public spending on schools must be specifically targeted to the
needs of the poor. This will apply to both the General Education (grades 1-9) and
the Further Education and Training (grades 10-12) phases" (Our emphasis).
xiv
_______________________________________________________________________________________
REFERENCES
Adedian I, Ajam T, and Walker L (1997) Promises, Plans and Priorities: South
Africa's Emerging Fiscal Structures. Cape Town: IDASA
Barberton, C (2002a) Comments on Chapter 14 of the Draft Consolidated
Report of the Committee of Inquiry into a Comprehensive System of Social
Security for South Africa: An occasional paper for the Budget Information
Service of Idasa. Cape Town: Idasa
Barberton, C (2002b) "South Africa" in Folscher, A (ed.) (2002) Budget
Transparency and Participation: Five African Case Studies. Cape Town:
Idasa
Blecher, M and Thomas, S (2003) "Health Care Financing" in South African
Health Review 2003: 269-290
Bolowana, A (2004) "Many Skip Pre-Primary Because of Fees" in The Mercury,
21 October 2004
Brand, D (2004) "Budgeting and Service Delivery in Programmes Targeted at the
Child's Right to Basic Education" in Coetzee E and Streak J (eds.) (2004)
Monitoring Child Socio-Economic Rights in South Africa: Achievements
and Challenges. Cape Town: IDASA
Cabinet
Statement,
18
September
gcis.gov.za/media/cabinet/020918.htm
2002:
http://www.
Cabinet
Statement,
20
September
gcis.gov.za/media/cabinet/020920.htm
2002:
http://www.
Cape Times (2005) "I Million Children to Go Hungry After Feeding Cuts" in Cape
Times, 28 September 2005
Child Health Unit (1997) An Evaluation of South Africa's Primary School
Nutrition Programme. Durban: Health Systems Trust
43
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Department of Education (2003a) Report to the Minister: A Review of the
Financing, Resourcing and Costs of Education in Public Schools, 03 March
2003. Pretoria: Government Printers.
Department of Education (2003b) Plan of Action: Improving Access to Free
and Quality Basic Education For All, 14 June 2003. Pretoria: Government
Printers.
Department of Education (2004) Government Notice No 26911 of 2004: Call
for Comment on the Education Laws Amendment Bill, 2004. Pretoria:
Government Printers
Economic Policy Research Institute (2004) The Social and Economic Impact
of South Africa's Social Security System. A Report Commissioned by the
Economics and Finance Directorate, Department of Social Development, 30
September 2004
Hunter N, May J, and Padayachee V (2003) Lessons for PRSP from Poverty
Reduction Strategies in South Africa. A Paper Presented at the Third Meeting
of the African Learning Group on the Poverty Reduction Strategy Papers, 3-5
September, Addis Ababa, Ethiopia
Kloka, D (2003) Integrated Nutrition Programme: School Feeding.
Background Document for Cabinet Briefing. Pretoria: Department of Health
Louw R, Bekker E, and Wentzel-Viljoen E (2001) External Evaluation of
Certain Aspects of Primary School Feeding. A Report Submitted to the
Department of Health, March 2001
Makgetla, NS (1995) "The new privatisation debate" in South African Labour
Bulletin 19(1): 65-73
Martin, P (2002) "Cabinet Stumbling on Crucial Steps to Curb Poverty" in The
Star, 10 December 2002
Mtyala, Q (2003) "School Kids Make a Meal of Food Plan" in Cape Argus, 07
May 2003
Naidoo, S (2003) "Asmal Confers Over Funding Method" in Business Day, 21
May 2003
National Treasury (1998) Division of Revenue Act. Pretoria: government
Printers
National Treasury (2000) Budget Review 2000. Pretoria: Government Printers
National Treasury (2001a) Intergovernmental Fiscal Review 2001. Pretoria:
Government Printers
National Treasury (2001b) Division
Government Printers
of
Revenue
Act
2001.
Pretoria:
44
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
National Treasury (2003) Intergovernmental Fiscal Review 2003. Pretoria:
Government Printers
National Treasury (2003) Budget Review 2003. Pretoria: Government Printers
National Treasury (2005a) Intergovernmental Fiscal Review 2005. Pretoria:
Government Printers
National Treasury (2005b) Budget Review 2005. Pretoria: Government Printers
Republic of South Africa (1987) White Paper on Privatisation and
Deregulation in the Republic of South Africa. Pretoria: Government Printers
Segal, S (1996) "Heat is Turned on in the RDP Office" in Mail and Guardian, 02
February 1996
Sidley, P (1995) "The Budget Social Upliftment is Given a Kick-start" in Mail and
Guardian, 17 March 1995
South African Reserve Bank (1996) "Annual Economic
http://www.resbank.co.za/economics/year1996/yearrep:html
Report
1996"
Steyn, N and Labadarios, D (2002) "Nutrition Policy Implementation" in South
African Health Review 2002: 327-349
Terreblanche, C (2002) "Cabinet to Discuss Impact of Food Inflation on Poor" in
Cape Times, 19 September 2002
The Office of the President (1994) White Paper on Reconstruction and
Development, 15 November 1994. Pretoria: Government Printers
Van der Walle (1998) "Targeting re-visited" in World Bank Research Observer
13(2): 231-248
Wildeman, RA (2000) Redistribution of School Funding. Budget Brief No. 48,
Cape Town, IDASA
Wildeman, RA (2001) School Funding Norms 2001: Are More Poor Learners
Benefiting? Budget Brief No. 79, Cape Town: IDASA
Wildeman, RA (2003) The Proposed New Funding in Provincial Education: A
Brave New World? An Occasional Paper for the Budget Information Service,
Idasa. Cape Town: Idasa
Wildeman, RA (2004) Reviewing Provincial Education Budgets 2004, Budget
Brief No. 143. Cape Town: Idasa
Wildeman, RA and Nomdo, C (2004) Implementation of Universal Access to
the Reception Year (Grade R): How Far Are We? An occasional paper for the
Budget Information Service of Idasa. Cape Town: Idasa
______________________________________________________________________________________________________________
45
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
APPENDIX A: WRITTEN CORRESPONDENCE WITH PROVINCIAL COORDINATORS OF THE SCHOOL NUTRITION PROGRAMME
Eastern Cape - Dr E. Fray, Acting Chief Director: Specialized Education
Services
Free State – Mr. H. R Khosa, Director: Learner Support
Gauteng – Ms. N. Rakwena, Project Manager: NSNP
KZN – Mr. N Mpanza, Director: Education; Dr. L.T Mbatha, Chief Director:
Curriculum Planning & Support; and Mr. M. Kubeka, Acting Manager: NSNP
Limpopo – Mr. M. Zitha, Provincial Manager: NSNP
Mpumalanga: Mr. J. Moya, Chief Education Specialist: NSNP
Northern Cape: Mr. B. Jacobs, Provincial Coordinator: NSNP
North West: Mr. K. I. Mpshe, Chief Education Specialist
Western Cape: Mr. D. Matunda, Programme Co-ordinator: NSNP
National Department of Health: Mr Jan Booysen provided important
historical spending and output data
National Department of Education: We received a short note from the
Department of Education through Ms Buyisiwe Ngidi
___________________________________________________________________________
APPENDIX B: LIST OF DATA TABLES
Table A1: Table 1: Household expenditure models of food shares
Predictor variables
ln (household income per capita)
All food items
Coefficient
P-value
Basic food items
Coefficient
P-value
-0.077619
0.000
-0.045139
0.000
Household pension amount
1.520000
0.000
0.869000
0.000
remittance received by household
0.000276
0.995
0.011400
0.674
-0.000178
0.000
0.000000
0.000
-0.000001
0.000
0.000000
0.000
0.368630
0.000
0.080860
0.074
(household pension amount)Sq
(remittance received by household)
Sq
years of education attained by the
household head
46
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
(years of education attained by the
household head)Sq
child support grant
disability grant
number of children
number of adult non-eligible for
pension
number of male elderly
number of female elderly
-0.034800
1.474980
2.495010
-1.131300
0.000
0.002
0.000
0.000
-0.013520
1.177060
1.252380
-0.531490
0.000
0.000
0.000
0.000
-0.745570
0.960430
-0.902540
0.000
0.038
0.008
-0.441240
0.458000
-0.434370
0.000
0.119
0.043
age of the household head
-0.060260
0.000
-0.046580
0.000
(age of the household head)Sq
female household head
black/african
coloured
indian/asian
rural
remittance *black
pension amount*black
predicted income*black
remittance*poverty
pension amount *poverty
0.000057
-0.769440
2.567240
3.343110
0.597420
1.571730
0.000254
-0.000303
-0.000200
-0.000408
-0.001340
0.000
0.001
0.000
0.000
0.377
0.000
0.000
0.001
0.000
0.005
0.000
0.000042
-0.160630
1.985260
-0.117260
-0.653830
1.438900
0.000125
-0.000145
-0.000159
-0.000225
-0.000887
0.000
0.259
0.000
0.720
0.127
0.000
0.000
0.012
0.000
0.015
0.000
0.002130
-1.680700
-4.640740
-2.842830
-8.700380
0.318310
-5.475550
-2.145050
-5.370670
-5.527330
102.332300
0.000
0.011
0.000
0.000
0.000
0.418
0.000
0.000
0.000
0.000
0.000
0.001100
-0.826660
-0.777930
-3.008320
-4.056780
1.424990
-3.015200
-0.588320
-2.754130
-1.530720
59.111780
0.000
0.049
0.003
0.000
0.000
0.000
0.000
0.017
0.000
0.000
0.000
predicted income *poverty
poverty
eastern cape
northern cape
free state
kwazulu natal
north west
gauteng
mpumalanga
limpopo
Constant term
Source: Economic Policy Research Institute (2004: 78)
Table A2: Actual expenditure as a percentage of budgeted expenditure on the
school nutrition programme by provincial education department, 2004/05
Province
Eastern Cape
Free State
Gauteng
KwaZulu-Natal
Limpopo
Mpumalanga
Total available
(R thousand)
177259
49100
75730
181420
153125
64079
Total spent
(R thousand)
170910
46149
75148
148459
194672
62696
Spent as a % of
total available
96.4%
94.0%
99.2%
81.8%
127.1%
97.8%
47
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Northern Cape
North West
Western Cape
Total
22469
72401
36617
832200
21440
64150
36191
819815
95.4%
88.6%
98.8%
98.5%
Source: Personal communication with the national Department of Education 2005
Table A3: Estimated cost per learner in 2004/05 and 2005/06 by provincial
education department
Estimated
Estimated cost
cost per
per
learner, Estimated total cost in learner,
Estimated total
Province
2004/05
2004/05
2005/06
cost in 2005/06
Eastern Cape
R 1.11
R164 850 870
R1.16
R181 036 365
Free State
R 1.50
N/A
R1.50
R50 000 000
Gauteng
R 1.10
R357 539,60
R1.10
440 952.60
KwaZulu Natal
R 0.90
R187. 9 million
R0.90
R213. 5 million
Limpopo
R 1.00
R176 million
R1.10c
R207 million
Mpumalanga
R 0.83
R69,079,000.00
R0.91
R70,235,000.00
R 1.00 - feeding
Northern Cape
R0, 15 - admin
R18, 876, 000 for feeding only R1.00
R19, 413, 576
R1.50-Urban
995 88590.4
R2.34
R891180
North West
R1.30-Rural
Western Cape
R 1.13
R33 503 817.609
R1.24
R40 135 000
Source: Personal communication with provincial education departments in
2004 and 2005
Note: we could not verify the exact interpretation of the estimated total cost in 2004/05 and 2005/06 and made a
decision to omit this table from the main text. However, for completeness of information sake, we include it in the
appendix.
Table A4: Total number of school targeted and number of schools reached as a
proportion of targets, 1994/95 to 2003/04
Financial Year
Targeted
1994/95
1995/96
1996/97
1997/98
1998/99
1999/2000
2000/01
2001/02
2002/03
2003/04
Source: Personal communication with
and 2005
Reached
Proportion reached
0.83
0.79
0.77
0.81
0.90
0.96
0.96
0.96
0.95
0.95
15 911
13 167
20 110
15 894
17 025
13 061
17 945
14 549
17 500
15 776
16 087
15 428
16 200
15 600
16 600
16 000
17 000
16 107
17 000
16 200
Departments of Health and Education in 2004
48
Reviewing Ten Years of the School Nutrition Programme: The Budget Information
Service, IDASA
Table A5: Total number of learners targeted and number of learners reached as a
proportion of targets, 1994/95 to 2004/05
Financial Year
Targeted
1994/95
6 29 3000
Reached
5 62 8000
Proportion reached
0.89
1995/96
1996/97
6 877 000
6 075 000
5 567 000
4 880 000
0.81
0.80
1997/98
1998/99
6 024 000
5 574 000
5 021 000
4 83 0000
0.83
0.87
1999/2000
2000/01
5 422 000
4 950 000
4 719 000
4 470 000
0.87
0.90
2001/02
4 850 000
4 550 000
0.94
2002/03
4 807 997
4 567 597
0.95
2003/04
4 850 000
4 600 000
0.95
2004/05
4 659 571
4 977 586
1.07
Source: Personal communication with Departments of Health and Education in 2004
and 2005
49